AN ABSTRACT OF THE THESIS OF
JAMES FRANKLIN AMOS for the DOCTOR OF PHILOSOPHY (Name) (Degree) m FOOD SCIENCE presented on Uuiin \3,ff7i (Major) -$Date)
Title: THE INFLUENCE OF MOLD CONTAMINATION, PROCESS-
ING AND MATURITY ON THE VOLATILES OF THE
STRAWBERRY, FRAGARIA ANANASSA DUCH.
Abstract approved: Yt - ' M_ Dr. R. E. Wrolstad
The effects of mold level (Botrytis cinerea), processing method
(fresh, individually quick frozen (IQF), frozen sugared sliced (FSS), freeze dried (FD), canned, and preserves), and maturity (underripe, normal, and overripe) and strawberry volatiles were analyzed by gas liquid chromatography (GLC) using on-column entrainment of head- space volatiles in aqueous extracts. Computerized analyses were used to calculate peak areas and to perform analyses of variance comparing the area of each peak for different samples. Peaks vary- ing significantly (P < 0.05) with processing method, maturity, and mold count were collected on Porapak Q columns, transferred to GLC columns, and identified by coupled GLC-mass spectrometry and by
GLC retention times. All volatiles whose peak areas changed significantly with mold level (P < 0.05) decreased as mold level increased. These volatiles were as follows: n-propionic acid, n-butyric acid, acetaldehyde, isobutyraldehyde, methyl acetate, 1, 1-diethoxymethane, 1,1- diethoxyethane, acetophenone, propiophenone, 1-phenyl-1, 2- propanedione, diethyltoluene, and one unknown. Neither ethanol nor diacetyl changed significantly with mold level.
Volatiles increasing significantly on processing were found in canned fruit and preserves. These were dimethyl sulfide, benzalde- hyde, furfural, and 1-propanethiol. In canned fruit and preserves
5-(hydroxymethyl)-2-furfural was newly formed. Many compounds
(acetals, acids, alcohols, aldehydes, esters, ketones, and hydro- carbons) decreased significantly in level on processing; all were identified. Numbers of compounds decreasing significantly with each process were as follows: IQF, five lower-boiling compounds; FSS, two lower-boiling compounds; FD, 37 compounds; canned, eight compounds; and preserves, 29 compounds.
In general most compounds varying significantly (P < 0. 05) with maturity increased. No new compounds were formed on ripening, and no compounds completely disappeared with ripening. The compounds varying significantly with maturity were esters (methyl acetate; ethyl formate, acetate, propionate, and butyrate; n- pentyl _n-hexanoate; P - phenylethyl acetate; cis-3-hexen- 1-yl hexanoate; and benzylacetate), aldehydes (acetaldehyde, isobutyraldehyde, benzaldehyde, and ethyl- benzaldehyde), acetals (1, 1-dimethoxymethane, 1, 1-diethoxymethane,
1, 1-diethoxyethane, and 1, 1-diethoxyoctane), acids (propionic, _n.- butyric, and isobutyric), aromatic ketones (acetophenone, propio- phenone, propylphenyl ketone, and 3-phenylpropan-2-one), ethanol, and one unknown.
These results would not seem to encourage use of this technique for monitoring mold levels in strawberries. The Influence of Mold Contamination, Processing and Maturity on the Volatiles of the Strawberry, Fragaria ananassa Duch.
by
James Franklin Amos
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
June 1972 APPROVED:
Assistant Professor of Food Science and Technology in charge of major
f T^ PV Head of Department of Food Science and Technology
Dean of Graduate School
Date thesis is presented lUsHI Typed by Mary Jo Stratton for James Franklin Amos ACKNOWLEDGEMENT
I would like to express my sincere appreciation for the encourage- ment and guidance given me by Dr. R. E. Wrolstad in this investiga- tion and in the rest of my doctoral program.
Thanks are also due to the other members of my program committee, Dr. R. F. Cain, Dr. W. D. Loomis, Dr. A. W. Anderson, and Dr. K. E. Rowe, for their willing assistance.
The cooperation and helpful suggestions of the faculty and graduate students of the Department of Food Science and Technology are gratefully acknowledged. Special thanks are due Dr. L. M.
Libbey for his help in obtaining mass spectra and Dr. M. E. Morgan for his help with the technique for Porapak collection of volatiles.
Thanks are also expressed to Dr. P. H. Krumperman, Dr. R. A.
Scanlan, and Dr. M. W. Montgomery.
The assistance of the personnel of the Computer Center, particularly Dr. L. C. Hunter's providing of the grant for computer time, is very much appreciated.
The advice of Dr. A. D. Campbell, Food and Drug Administra- tion, was extremely helpful in planning the study.
I would like to express my appreciation to the U. S. Public
Health Service and the General Foods Corporation for their support of the work. TABLE OF CONTENTS
Page
INTRODUCTION 1
REVIEW OF LITERATURE 3
Strawberry Volatiles 3 Identification of Volatiles 3 Effect of Various Parameters on Volatile Composition 6 Microbially-Produced Volatiles 8 Use in Identification of Microorganisms 8 Use as a Food Quality Indicator 9 Mold-Produced Volatiles 10 Botrytis cinerea 12 Economic Significance 12 Growth Conditions 14 Products 17 Enzymes 17 Detection 19
EXPERIMENTAL 20
Strawberries 20 Source 20 Mold-Level Classifications 20 Maturity Classifications 21 Fruit for Processing 21 Processing Methods 22 Fresh 22 Individually Quick Frozen 22 Frozen Sugared Sliced 23 Freeze Dried 23 Canned 23 Preserves 24 Microbiological Materials and Methods 24 Mold Counts 24 American Type Culture Collection Culture 24 Mold Culture from Strawberries 25 Comparison of Mold Cultures 25 Analysis of Headspace Volatiles 26 On-Column Entrainment 2 6 Gas Liquid Chromatography 27 Page
Quantitation of GLC Data 28 Statistical Procedures 29 Identification of Volatiles 2 9 Porapak Collection 29 Mass Spectrometry 33 GLC Retention Times 34
RESULTS AND DISCUSSION 35
Effect of Mold Level 36 Mold Counts 36 Comparison of Mold Cultures 36 Comparison of Volatiles 38 Effect of Processing Method 48 Effect of Freezing (IQF and FSS) 53 Effect of Freeze Drying 58 Effect of Canning 5 9 Effect of the Preserving Process 62 Effect of Maturity 63 Esters 70 Aldehydes 71 Acetals 72 Acids 73 Aromatic Ketones 73 Alcohols 1 74 Other Considerations 75
BIBLIOGRAPHY 77
APPENDICES 92 LIST OF FIGURES
Figure Page
1 Headspace assembly for the entrainment of volatiles on the Porapak column. 31
2 Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a TRIS column, ('very moldy (VM) and normal good quality (NGQ) fruit). 39
3 Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a Carbowax column (very moldy (VM) and normal good quality (NGQ) fruit). 40
4 Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a TRIS column (freeze dried and fresh fruit). 49
5 Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a TRIS column (preserves and fresh fruit). 50
6 Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a Carbowax column (freeze dried and fresh fruit). 51
7 Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a Carbowax column (canned and fresh fruit). 52
8 Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a TRIS column (underripe (UR), normal good quality (NGQ) and overripe (OR) fruit). 64 Figure Page
9 Analysis of the headspace volatiles present in an aqueous extract o*. individually quick frozen (IQF) straw- berries using a Carbowax column 65 (underripe (UR), normal good quality (NGQ) and overripe (OR) fruit). LIST OF TABLES
Table Page
1 Howard mold counts for strawberries of different subjective mold-level classifica- tions processed in different ways. 37
2 Normalized peak areas of TRIS volatiles and comparison of different mold levels. 41
3 Normalized peak areas of Carbowax volatiles and comparison of different mold levels. 41A
4 GLC-MS identification of volatiles important in mold-level differences. 42
5 Comparison of relative amounts of TRIS compounds decreasing on mold growth with relative amounts of the products to which they could have been altered. 46
6 Normalized peak areas of TRIS volatiles and comparison of processed with fresh fruit. 54
7 Normalized peak areas of Carbowax volatiles and comparison of processed with fresh fruit. 55
8 GLC-MS identities of TRIS volatiles differing in amount with processing method. 56
9 GLC-MS identities of Carbowax volatiles differing in amount with processing method. 57
10 Normalized peak areas of TRIS volatiles and comparison of different maturities. 66
11 Normalized peak areas of Carbowax volatiles and comparison of different maturities. 67 Table Page
12 GLC-MS identities of TRIS peaks varying with maturity. 68
13 GLC-MS identities of Carbowax peaks varying with maturity. 69 THE INFLUENCE OF MOLD CONTAMINATION, PROCESSING AND MATURITY ON THE VOLATILES OF THE STRAWBERRY, FRAGARIA ANANASSA DUCH.
INTRODUCTION
Molds have long been known to be responsible for economic loss and to cause infectious diseases; their presence in most foods is directly related to spoilage and decomposition (Wilson, 1966; Anony- mous, 1970). Administrative guidelines for mold levels in different food products have been established by the Food and Drug Administra- tion (Anonymous, 1970),, Mold levels in excess of these values are considered to represent violation of good manufacturing practices, as they are indicative of the use of excessively decomposed product or of unsanitary processing practices (ibid. )„ With the current recognition that molds play a significant role in the toxic diseases of man because of their elaboration of mycotoxins, the problem of analytical mycology of food assumes a role of added importance (Wilson, 1966; Eisenberg,
1969).
The Howard mold count procedure, though simple and rapid, is still somewhat subjective (Giannone and Tommasicchio, 1969).
Furthermore, it is useless for detection of the inclusion of moldy fruit in products where the solid mass of the fruit is no longer present, as in the case of juices, jellies, and flavor essences (Campbell, 1970).
Fields et al. (1968) and the Food and Drug Administration (as summarized by Salwin, 1970) have conducted some research in the area of chemical indices of microbial quality. However, these studies have not delved deeply into the detection of a mold by its alteration of the pattern of volatiles of an actual food product.
The primary objective of this study was to use gas liquid chromatography (GLC), mass spectrometry (MS), and statistical analyses to compare volatiles of good quality strawberries with those of moldy fruit to see if chemical indices of mold growth on straw- berries could be found. It was desired to make this comparison for fruit prepared by several normal industrial processes in order to test the applicability of the method. A secondary objective was to use these same analytical and statistical procedures to examine the volatiles of strawberries processed in different ways and the volatiles of strawberries of different maturities in order to see what informa- tion could be gleaned in these little-investigated aspects of the nature of the volatile components of strawberries. REVIEW OF LITERATURE
Strawberry Volatiles
Identification of Volatiles
The volatile constituents of strawberries have received much attention in Europe and in the United States. To date nearly 150 components have been isolated. These have been reviewed by Nursten and Williams (1967) and by Nursten (1970).
Investigation of strawberry aroma began with Coppens and
Hoejenbos (1939). Using steam distillation, fractional distillation, melting and boiling point determinations, and various chemical reactions they found ethyl acetate, ethyl butyrate, a hexanoic ester, isoamyl esters, borneol, terpineol, isofenchyl alcohol, and several unidentified compounds.
Winter and his colleagues in Switzerland used a special distilla- tion apparatus which crushed and distilled the fruit under nitrogen
(Winter et al. , 1958). They then formed derivatives and identified compounds by paper chromatography (Winter, 1958; Seidel et al. ,
1958). Compounds identified'were; cis-3-hexen- l-ol, 1-hexanol, ethanol, methanol, 3-methyl-3-buten- l-ol, 1-penten-3-ol, 1-pentanol, trans-2-buten- l-ol, 1-butanol, 3-methyl-3-buten-l-ol, and geraniol
(Winter and Sundt, 1962); propanol, acrolein, crotonal, 2-pentenal, 4 cis-3-hexenal, 2-pentanone, l-penten-3-ol, 2-heptanol, benzyl
alcohol, propyl acetate, methyl hexanoate, hexyl butyrate, isopropyl
butyrate, trans-2-butenoic acid, and trans-2-hexenyl-1-yl-butyrate
(Winter and Willhalm, 1964); and formic acid, acetic acid, propionic
acid, n-butyric acid, isobutyric acid, hexanoic acid, methyl formate,
ethyl formate, methyl propionate, ethyl propionate, methyl isobutyr-
ate, and ethyl isobutyrate (Willhalm £t al. , 1966).
In early efforts by the U.S. Department of Agriculture's
Western Utilization Research and Development Division (Albany,
Calif. ) several compounds were found by Dimick and his co-workers.
They separated and identified the following volatiles using aqueous
distillation, GLC, and derivative formation: 2-hexenal, acetone, dlacetyl, ethanol, methanol, several unidentified esters, _n-decanoic
acid, n-pentanoic acid, n-butyric acid, isobutyric acid, and acetic
acid (Dimick and Makower, 1956; Dimick and Corse, 1958; Corse and
Dimick, 1958).
In later efforts at the same laboratory, Teranishi and co-
workers used temperature-programmed capillary column GLC with
flame ionization detection, infrared spectroscopy, nuclear magnetic
resonance, and MS to identify more than 100 volatiles from straw-
berries. In the publication by Teranishi et al. (1963) they reported
isopentane, methylpentane, diethyl ether, ni-hexane, acetaldehyde,
methylcyclopentane, 2-methyl- 1-pentene, cyclohexane, acetone, 5 1, 1-dimethoxymethane, 1, 1- dimethoxyethane, methyl acetate, 1,1- methoxyethoxyethane, ethyl acetate, 1, 1-diethoxyethane, methyl iso- butyrate, 3-methyl-2-butanone, benzene, 1, 1-ethoxypropoxyethane, ethyl propionate, ethyl isobutyrate, methyl n-butyrate, and several unidentified compounds. In the publication by McFadden et_ al. (1965) they reported the following additional compounds: ethanol, 1, 1- diethoxypropane, 1,1-methoxybutoxyethane, Z-methylbutan-2-ol, methyl Q'-methylbutanoate, isoamyl formate, amyl formate, 1, 1- ethoxybutoxyethane, ethyl n-butyrate, ethyl a-methylbutanoate, n- butyl acetate, 1, 1-diethoxybutane, 2-hydroxy-3-methylbutane, ethyl isovalerate, 1, 1-methoxypentoxyethane, 2-hexyl acetate, Z-pent&nol,
1, 1-diethoxypentane, ji-butanol, isoamyl acetate, 1, 1-ethoxy- pentoxyethane, 3-methyl-1-butanol, 2-methyl-1-butanol, methyl n- hexanoate, 2-hexenol, ethyl ^-hexanoate, 1, 1-methoxyhexoxyethane, isopropyl hexanoate, 1, 1-diethoxyhexane, ji-hexyl acetate, 1, 1- ethoxyhexoxyethane, ethyl heptanoate, hexenyl acetate, 1, 1-ethoxyhex-
3-enoxyethane, n-hexanol, ethyl heptanoate, 3-hexene-l-ol, 2-hfexene- l-ol, butyl hexanoate, furfural, methyl octanoate, ji-butyl n- hexanoate, _n-hexyl n-butyrate, 2-acetylfuran, ethyl octanoate, butyl a-methylbutyrate, 1, 1-diethoxypentane, benzaldehyde, pentyl hexanoate, methylfurfural, linalool, pentenyl hexanoate, acetophenone, methyl _n- decanoate, ethyl benzoate, n-hexyl n-hexanoate, 1, 1-di-n- hexoxyethane, cis-3-hexen-1 -yl hexanoate, ethyl _n-decanoate, trans-3-hexen- 1-yl hexanoate, benzyl acetate, a-terpineol, 2-
hexen-1-yl hexanoate, pentyl octanoate, naphthalene, 1,1-
diethoxyoctane, (3 -phenylethyl acetate, 1, 1 - dihexenoxyethane, 2-
methylnaphthalene, hexyl octanoate, 1-methylnaphthalene, ethyl
dodecanoate, hexenyl octanoate, cis-ethyl cinnamate, trans-methyl
cinnamate, and trans-ethyl cinnamate.
Further investigations were carried out by Tressl^t ah (1969).
They confirmed the identification of compounds reported earlier by
other workers but failed to identify any new compounds.
The compounds found by different workers were summarized by
Nursten (1970). By class their numbers are as follows: acids, 12;
alcohols, 27; esters, 52; carbonyls, 21; acetals, 20; and hydrocar-
bons, 10. This represents a total of 142 compounds.
Effect of Various Parameters on Volatile Composition
Varietal differences in the volatiles of strawberries have
received little attention. Black et al. (1971) compared the volatiles of
four different varieties of strawberries. They found minor quantita- tive differences but no qualitative differences.
Several investigators have studied the influence of processing method on strawberry volatiles. Dimick and Makower (1956) reported
formation of n-hexanal from 2-hexenal during one year of frozen 7
storage. They also found that heating \of strawberry puree resulted in
the loss of lower boiling volatiles. Katayama et al. (1968) published
information on severe loss of volatiles in the processing of strawberry
jam. Workers at the Sprenger Institute (1969) substantiated the finding
of Dimick and Makower (1956) that n-hexanal was formed from 2-
hexenal during one year of frozen storage and proposed that formation
of n-hexanal accounted for the off-flavor associated with frozen straw-
berries. They found that added sucrose inhibited this off-flavor
formation. Sloan et al. (1969) found that acetaldehyde increased and
dimethyl sulfide, isobutyraldehyde, furan, furfural, 2-acetyl furan,
and ethyl furoate were produced when pureed strawberries were heated
to 1200C for 30 min.
The only reported study on the effect of maturity on strawberry
volatiles was done by Ahmed and Scott (1963). Using GLC they found
seven peaks in unripe Fairfax strawberries as compared to 12 peaks
in ripe fruit of the same variety. For Dixieland strawberries they
found eight peaks in unripe fruit and 12 peaks in ripe fruit. For both
varieties peaks present in unripe and ripe fruit increased in the ripe
fruit. No peaks were identified by these 'workers.
No -work has been reported previously on the effect of mold
growth per se on strawberry volatiles. Sandoval and Salwin (1968)
did use GLC to compare the volatiles of good quality strawberries with those of strawberries decomposed by 10 days storage at 10 C. 8 Differences in the chromatograms were observed, but no compounds were identified.
Microbially-Produced Volatiles
Use in Identification of Microorganisms
Gas liquid chromatography has been used in several cases to identify microorganisms by their patterns of volatile compounds.
Alexander (1970) stated that using GL.C, one population could be recognized in a collection of different types of organisms. He also asserted that a smaller population could be detected using GLC than using any other method found heretofore. This was stated to be particularly true when an electron capture detector was employed.
Henis et al. (1966) found that they could differentiate each of 29 strains of bacteria based on the relative sizes of the peaks in their
GLC patterns of volatiles. Reiner (1965) successfully differentiated
95 strains of bacteria based on gas liquid chromatograms of the pyrolysis products of a sample of lyophilized whole bacteria.
Gas liquid chromatography has been used to detect levels of viral infection. Serum samples were used to determine the activity of equine infectious anemia virus (Mitruka et^ al. , 1968) and to detect the growth of the infectious hepatitis virus in dogs (Mitruka and
Alexander, 1968). 9 Bassette et al. (1967) used GLC to identify six different genera
of bacteria in milk. Guarino and Kramer (1969) stated that GLC
analysis of headspace vapors over cultures of bacteria in foods was
reproducible enough to be used as a tool for routine microbiological
assay.
Use as a Food Quality Indicator
The growth and action of microorganisms is known to have an
important effect on food flavors. For example, the characteristic
culture flavor of butter, buttermilk, sour cream, and creamed cottage
cheese is primarily dependent on the flavor and aroma substances
produced by lactic-acid producing bacteria (Lindsay, 1967).
Angelini and Pflug (1967) proposed that unidentified micro-
organisms growing on apples in a controlled atmosphere storage room
might have produced some of the oxygenated compounds (mainly
alcohols and esters) isolated in the atmosphere of the chamber.
Kretovich and Prokhovora (i960) found that spoilage of grain was
accompanied by an increase in volatiles, mainly aldehydes, ammonia,
and amines.
The Food and Drug Administration and other regulatory
agencies have investigated GLC of volatiles as an index of decomposi- tion and filth in foods. Salwin (1970) reported the use of a method for
quantitative determination of lactic and succinic acids as indicators of 10
decomposition in eggs. Moorhouse and Salwin (1969) found that the
determination of volatile reducing substances was valid for assaying
decomposition in raw foods and most cooked foods but that this test was no longer useful after freeze drying. Sandoval and Salwin (1968),
in their study of the volatiles of good quality and decomposed straw- berries, found that differences in the chromatograms were still apparent after freeze drying. Salwin (1969) used diacetyl, acetyl- methylcarbinol, ethanol, and acetic acid as indicators of decomposi- tion in apple products. Strachenbrock (1962) reported that ethanol was used in Germany as an indicator of deterioration of fruit juices.
Production of free amino acids was used by Rohan and Stewart
(1967) to assess the degree of fermentation of cocoa beans. Fields jet al. . (1968) investigated acetylmethylcarbinol, diacetyl, 2,3- butyleneglycol, volatile acids, nonvolatile acids, ethanol, and tyrosine as indicators of microbial spoilage in various food products. They found that diacetyl could be used as a presumptive indicator of sanita- tion in an apple juice processing plant. Giannone and Tommasicchio
(1969) searched for a chemical method based on change in volatile compounds to replace the Howard mold count in tomato products.
They could find no correlation between mold counts and chemical data.
Mold-Produced Volatiles
The identification of specific volatiles produced by molds has 11 received only a small amount of attention. Romano and Safferman
(1963) found a compound produced by certain actinomycete cultures
(Streptomyces griseoluteus and S. odorifer) which was responsible
for the musty odor of damp cellars, aged straw piles, and stagnant
ponds. Gerber and Lechevalier (1965) studied the infrared and ultra-
violet spectra of the compound. Dougherty et al. (1966) isolated the
compound and proposed several structures for it. Using retention times on different GLC columns Rosen et al. (1968) identified this
compound as geosmin. Medsker et al. (1968) investigated the
structure of the compound using MS, infrared spectroscopy, and
proton magnetic resonance. They characterized geosmin (C _H__0)
as a dimethyl substituted, saturated, two-ring tertiary alcohol with the hydroxyl group very sterically hindered.
Kaminski et al. (1969) investigated the origin of raustiness in grain and attempted to identify the compounds responsible. They found and isolated two compounds produced by Aspergillus flavus growing on grain, one a compound with a musty odor and the other a
compound with a mushroom-like odor. These compounds were later
'identified by mass spectrometry (Libbey, 1971). The musty compound was found to be oct-2-en- l-ol, and the mushroom-like compound was found to be oct-l-en-3-ol.
Fields and Richmond (1967) investigated the effect of different pH's (4. 0 and 7. 0) on the production of fungal metabolites by Botrytis 12
cinerea, Rhizopus nigricans, a Mucor sp. , Alternatia tenuis, and
Rhizoctonia solani. The metabolites examined were acetylmethyl-
carbinol, ethanol, volatile acids, and nonvolatile acids. Nonvolatile
acid production (expressed as lactic) proved to be the most constant with different organisms and different pH conditions, thus showing the
greatest promise as a single indicator of the fungi tested.
Botrytis cinerea
13. cinerea Pers. is a mold which may appear gray, gray- green, brown, or brown-black (Gilman, 1957). It bears conidia on erect conidiophores (ibid. ). It may be classified taxonomically as follows (ibid. ):
Kingdom Planta
Subkingdom Thallophyta
Class Fungi imperfecti
Order Moniliales
Family Moniliaceae
Economic Significance
Botrytis cinerea is responsible for the most serious pathological condition in strawberries (Harvey and Pentzer, 1966). It has been said that this organism often causes loss of 10% or more of a straw- berry crop in the field (ibid. ). Powelson (I960) found this organism 13
to be the most important fungus causing fruit rot in strawberries in
the Pacific Northwest. Minor amounts of rot of strawberries were
caused by Rhizopus nigricans, Rhizoctonia spp. , Dendrophoma
obscurans, Gnomonia fructicola, and a Melancoluim sp. (ibid. ).
Botrytis cinerea has been found to overwinter in mummified
fruits and plant debris in strawberry fields (ibid. ) and to be carried
to the fruit by air currents or by contact (Harvey and Pentzer, 1966).
Powelson (I960) found that rot originated at the stem end of the fruit
and that there was latent infection in that area of a high percent of
marketable strawberries.
Motoc and Brad (1963) proposed that the phytopathologic activity
of B. cinerea was related to its highly active pectolytic enzymes.
Salkova et al. (1966) asserted that the capability of B. cinerea to rot
fruits was attributable partly to its production of pectinesterase and
partly to its production of polygalacturonase.
Powelson (I960) found that the rate of development of B^ cinerea
rot of strawberries depended on the degree of infection. A high level
of nitrogen fertilization increased the incidence of fruit rot (ibid. ).
Strawberry varieties differed in their susceptibility to B^. cinerea under
field conditions; Northwest was one of the least resistant varieties
(ibid. ). Rubin and Ivanova (1958) found that the ability of cabbage varieties to resist B. cinerea attack was related to the ability of the
plant to produce amino acid oxidases. The more resistant varieties 14 produced greater quantities of these enzymes than did the more susceptible varieties.
Infection of fruits by B. cinerea has not been deleterious in all cases. Sweet table wines (such as Sauternes) have long been made from grapes on which this mold has grown (Gray, 1959). The mold has been grown in pure culture, and the mycelium has been used to alter the flavor of white wines (King ^t ai. , 1969). Chaudhary et al.
(1968) compared the GL.C patterns of volatiles for normal and
Botrytized grapes. They found some variations in the relative amounts of individual components but were unsure as to whether these variations were great enough to account for the aroma differences between the two wines. Nelson and Amerine (1956) reported that wines from grapes inoculated with B. cinerea had a distinct, characteristic flavor, but the compound(s) responsible for the effect were not identified. De Jong et al. (1968) speculated that the effect of the growth of B_. cinerea on the flavor of white wines might be due to changes in oxidation-reduction equilibria brought about by oxidative enzymatic reactions.
Growth Conditions
The growth conditions for B_. cinerea have been thoroughly investigated. Mishra (1953) found that the organism required peptides and could not grow using simple amino acids as a nitrogen source. He 15 reported that a combination of maltose and yeast extract represented the best medium tried for growing the organism. King et al. (1969) reported maltose to be the best carbon source of those they examined for the organism. Other carbon sources which have been reported to be utilized by B. cinerea are glucose (Usseglio-Tomasset, 1958;
King et al. , 1969); sucrose (King e* al. , 1969); tartaric acid (Novak and Voras-Felaki, 1958; King et al. , 1969); succinic acid, citric acid, lactic acid, and sodium acetate (Novak and Voros-Felaki, 1958); ethanol (Novak and Voros-Felaki, 1958; Usseglio-Tomasset, 1958); volatile acids (Charpentie, 1954; Gray, 1959; Amerine and Cruess,
I960; Minarik, 196l); acetic acid, formic acid, and propionic acid
(Novak and Voros-Felaki, 1958); lower molecular weight acetals
(Popova and Puchkova, 1953); aromatic carbonyls (Henderson and
Farmer, 1955); sodium lactate, ascorbic acid, malonic acid, oxalic acid, and sodium oxalate (Novak and Voros-Felaki, 1958); and_n- hexanol, gamma-butyrolactone, and butylene glycol (Chaudhary et al. ,
1968). Harvey and Pentzer (1966) reported that the organism grew on dead organic material in strawberry fields. Mishra (1953) found that the trace elements zinc, copper, and manganese were necessary for gro'wth. King et al. (1969) found a carbonsnitrogen ratio approximately equal to that normally found in fruits caused the organism to develop conidia and a moldy flavor, while a carbon:nitrogen ratio lower than that found in fruits favored the production of mycelia and a mushroom- like flavor. 16
Harvey and Pentzer reported that B_. cinerea was a more common strawberry pathogen in cool producing regions that in warm ones. The organism was said to be able to grow at temperatures down to 15 C (Tanner, 1944). Two different optimum growth temperatures have been reported; 15 to 18 C (Cochrane, 1958) and 30 C (King et al. , 1969). The optimum, temperature for spore germination was mentioned to be 20 C (Harvey and Pentzer, 1966). King et al. (1969) found that mycelium yields were greatest at the lower temperatures within the 15 to 30 C range. The fungus would be killed by exposure to temperatures between -15 and -22 C, according to Cochrane (1958).
King et al. (1969) reported that IS. cinerea grew over the pH range from 3 to 8. They found that the greatest yield of mycelium was produced at pH 4 to 5.
According to Harvey and Pentzer (1966) growth of the mold was favored by a microclimate having a high relative humidity. Such an environment could be provided by heavy leaf cover or by debris surrounding the strawberry plants (ibid. ). Incidence of B_„ cinerea on strawberries was highly correlated (P < 0. 05) with high rainfall and high relative humidity (Jarvis, 1964). Tanner (1944) reported that
B. cinerea has a competitive advantage over most other fungi when low temperatures (below 20 C) are prevalent. Cochrane (1958) asserted that B_. cinerea germinates best at a relative humidity between 90 and 95% and does not require liquid water for germination. 17
Products
Botrytis cinerea was not among the fungi which were examined
for production of a mycotoxin (Gray, 1970). However, various other
products of B. cinerea metabolism have been investigated.
Several organic acids have been found to be produced by 13.
cinerea. Those which have been reported are as follows; oxalic acid
(Chaudhary et aL , 1968), malic acid (ibid. ), tartaric acid (ibid. ),
glyoxylic acid (Birkinshaw et al. , 1942), and gluconic acid (Usseglio-
Tomasset, 1958; Amerine and Cruess, I960). Usseglio-Tomasset
(1958) reported that the organism did not produce citric acid or
volatile acids.
Other compounds found to be produced by the organism are
glycerol (Usseglio-Tomasset, 1958; Amerine and Cruess, I960;
Dittrick, 1964), unidentified sugars (Amerine and Cruess, I960;
De Jong et al. , 1968), diethyl succinate (Chaudhary et al. , 1968), thiourea (Ovcharov, 1937), andm-cresol (Chaudhary et al. , 1968).
Fields et al. (1968) reported that the organism did not produce
significant quantities of diacetyl or ethanol.
Enzymes
Much work has been done in elucidating the enzymes of B_. cinerea, particularly the pectic enzymes. It has been found to 18 produce protopectinase (Cochrane, 1958), a very active endopoly- galacturonase (Hancock et al. , 1964), an exopolygalacturonase (ibid. ), unspecified polygalacturonases (Winstead and Walker, 1954;
Cochrane, 1958; Barashetal., 1964; Tagawa and Kaji, 1967), poly- methylgalacturonase (Hancock et al. , 1964), and pectinmethylesterase
(Winstead and Walker, 1954; Cochrane, 1958; Amerine and Cruess,
1960; Tagawa and Kaji, 1967). Van Den Berg and Tang (1969) reported that B. cinerea produced pectolytic enzymes in significant quantities only when readily metabolizable sources of energy such as sugars were not available. Botrytis cinerea has also been found to produce (3 -galactosidase (Pitt, 1968), (3-glucuronidase (ibid. ), cellobiase (Barashetal., 1964), xylanase (ibid. ), and cellulase
(ibid. ).
Several dehydrogenases have been reported. These are glucose-6-phosphate;NADP dehydrogenase (De Jong et al. , 1968), isocitrate dehydrogenase (Rubin and Ivanova, 1958), malate dehydro- genase (ibid. ), glutamate dehydrogenase (ibid. ), and alcohol dehydro- genase (ibid. ).
Pitt (1968) found that the organism produced an esterase. Franke et al. (196l) reported two transaminases, an alaninesglyoxylate transaminase and a glutamate;glyoxylate transaminase. The U.S.
Department of Agriculture (1968) found a protease produced by B. cinerea. Miscellaneous other enzymes reported include an acid 19 phosphatase (Pitt, 1968), an aryl sulfatase (ibid. ), and a phenol oxidase (De Jong et al. , 1968).
Detection
Only one study has been done heretofore concerning detection of
B. cinerea by use of its metabolites; this was reported by Fields and
Richmond (1967) and more completely by Fields et al. (1968). They found that the organism produced only a small amount of volatile acids at pH 7. 0 and essentially none at pH 4. 0. It produced insig- nificant amounts of acetylmethylcarbinol, ethanol, and diacetyl. Non- volatile acids (as lactic) were moderately good indicators of B. cinerea growth. Overall they concluded that the organism was a weak fermenter and was thus hard to detect by chemical means. 20
EXPERIMENTAL,
Strawberries
Source
The strawberries (Fragaria ananassa Duch„ variety Northwest) used in the mo^d-leVel and processing phases of this study were obtained from Oregon State University's North Willamette Experiment
Station, Aurora, Ore., on June 19 and 23, 1970. The strawberries
(Northwest variety) used in the maturity phase were obtained from the
Somogyi Berry Farm, Hood River, Ore., on June 30, 1970. All strawberries were washed in a commercial agitating washer before separation into experimental groupings.
Mold-Level Classifications
Fruit to be used for studying the effect of mold level on straw- berry volatiles was sorted subjectively into the following groupings:
1. Normal, good quality (NGQ) - Firm, red, ripe straw-
berries, undamaged by bruises or mold growth;
2. Fair quality moldy (FQM) - Strawberries with less than
one-half of their surface area covered by mold and with the
rest of the fruit firm red ripe;
3. Moldy (M) - Strawberries with greater than three-fourths 21
of their surface area covered with mold but without
appreciable loss of textural properties;
4. Very moldy (VM) - Strawberries with their surface area
covered with mold growth and with appreciable textural
breakdown.
Maturity Classifications
Fruit to be used for studying the effect of maturity on straw- berry volatiles was sorted subjectively as follows:
1. Normal good quality (NGQ) - Firm, red, ripe, strawberries
undamaged by mold growth or bruises;
2. Underripe (UK.) - Strawberries with normal red color on
one-fourth to three-fourths of their surface area and with-
out mold growth or bruises;
3. Overripe (OR) - Strawberries with moderate softening and
darker red coloration than NGQ fruit but without mold
growth or bruises.
Fruit for Processing
Fruit to be used for studying the effect of processing method on strawberry volatiles was treated in the following ways: individually quick frozen (IQF), frozen sugared sliced (FSS), freeze dried (FD), canned, and preserves. Fresh fruit was also examined. All straw- berries used in this phase of the study were NGQ. 22
Strawberries of different mold-level classifications were processed by different methods; those used were as follows:
NGQ - Fresh, IQF, FSS, FD, canned, preserves;
FQM - IQF, FD, canned;
M - Fresh, IQF, FSS, FD, canned, preserves;
VM - IQF.
Fruit used in the maturity phase of the study was IQF processed.
All three maturity classifications (UR, NGQ, and OR) were treated in this way.
Processing Methods
Fresh
On the same day it was taken from the field, fruit was washed and sorted, and a filtrate (see On-Column Entrainment section) was prepared. The volatiles present therein were analyzed by headspace on-column entrainment GLC (HSGL.C).
Individually Quick Frozen
Fruit was frozen overnight in an air blast freezer at -35 C.
Frozen fruit was placed in a polyethylene bag, sealed with a wire closure, and put in a 30 lb fruit tin for storage at -24 C. 23
Frozen Sugared Sliced
Fruit was sliced (1/2 in. slices) in an U~schel rotary blade slicer, mixed with sugar (four weights of berries; one weight of sugar) in a drum-type mixer, packaged in 16 oz fiber cartons, and frozen overnight in a -35 C air blast. The frozen fruit was stored at
-240C.
Freeze Dried
Whole fruit (IQF) was freeze dried in a Hull freeze dryer (Model
651 M-9-WDF20). The pressure ranged from 300 microns to 900 microns, and the maximum product temperature was 49 C. The dried product weighed 10% of its fresh weight. Freeze dried strawberries were stored under nitrogen in sealed 603 x 700 tin cans. A small packet of drying agent (silica gel) was enclosed in each can. Cans of freeze dried fruit were stored at 20 C
Canned
Whole strawberries were canned in 307 x 409 fruit-enameled tin cans with fruit-enameled ends. The fruit was covered with water at
96 C so that the can contents before exhausting were 90% fruit and 10% water. Unsealed cans were exhausted 8 min in a steam exhaust box.
Cans were sealed and cooked 8 min in boiling water at atmospheric pressure, cooled, and stored at 20 C. 24
Preserves
A standard vacuum pan process was used (Sunkist Growers, 1964;
Varseveld and Beavers, 1966). The final soluble solids content of the
preserves was 68 Brix. The preserves were sealed hot in 303 x 406
fruit-enameled tin cans with fruit-enameled ends, cooled, and stored
at 20OC.
Microbiological Materials and Methods
Mold Counts
Howard mold counts were done by the Consumer and Marketing
Service, U. S. Department of Agriculture, Salem, Ore. The method
used was that described in Horwitz (1965). Mold count results were
expressed as number of positive fields per 50 fields. Each sample was
counted in duplicate. Samples were counted as presented to the
agency, except that freeze dried samples were rehydrated. Mold
counts were normalized to give the number of positive fields per 50 fields in terms of unprocessed strawberries.
American Type Culture Collection Culture
The culture of Botrytis cinerea was obtained from the American
Type Culture Collection (A. T. C. C. culture number 11542, isolated by
F. A. Weiss). The culture was maintained on Saboraud maltose agar 25 (Difco Laboratories, 1953) and was kept at 4 C. The culture was transferred once a year, as recommended by the A. T. C. C.
Mold Culture from Strawberries
Cultures from VM strawberries were prepared on Saboraud maltose agar plates. Cultures were maintained on Saboraud maltose agar at 4 C and were transferred once a year.
Comparison of Mold Cultures
A filtrate (see On-Column Entrainment section) of NGQ straw- berries was used as a substrate for comparative studies on the effect of different mold cultures on strawberry volatiles. Samples were incubated in 300 ml baffled Erlenmeyer flasks in a Psychrotherm
Incubator Shaker at 30 C The samples were shaken and incubated for
144 hr. The following samples were used:
1. Strawberry filtrate (100 ml) plus 10 ml of a suspension of
the A. T. C. C. culture of B. cinerea in distilled water;
2. Strawberry filtrate (100 ml) plus 10 ml of a suspension of a
culture of the mold isolated from VM strawberries;
3. Strawberry filtrate (100 ml) plus 10 ml of distilled water.
After incubation duplicates of each sample were examined directly using HSGLC. 26
Analysis of Headspace Volatiles
On-Column Entrainment
All comparisons of the volatiles of different samples in the mold-
level, processing, and maturity phases of the study were done using
filtrates. Samples were pureed for 1 min in an Osterizer blender with the speed controlled by a rheostat with a setting of 50. The puree was filtered through four layers of cheesecloth, and the volume of filtrate was recorded so that the GLC peak areas for each sample
could be normalized for the amount of filtrate obtained.
The volatile compounds of different samples were compared by
HSGL.C using the gas entrainment on-column trapping technique
developed by Morgan and Day (1965) and modified by Heatherbell ^t al.
(1970). A 20 ml Kimble vial containing 10 ml of filtrate and an
internal standard (0. 5 ppm of 2-butanone for analyses on TRIS columns and 0. 1 ppm of n-heptanol for those on Carbowax columns) was used.
A few mg of 1-tetradecanol were added to prevent foaming. Samples were agitated by a magnetic stirrer and heated by a water bath.
Entrainment conditions for the TRIS column were;
Water bath temperature; 60 C - 1 ;
Entrainment time; 10 min.
Carbowax conditions were;
Water bath temperature; 65 O C -L- 1 o ; 27
Entrainment times 20 min.
The nitrogen flow rate was 25 ml/min for both columns. The cold bath was crushed dry ice in methyl cellosolve (-78 C). All samples were done in duplicate on the TRIS column and on the Carbowax column.
Gas Liquid Chromatography
The gas liquid chromatograph used was a Varian Aerograph
Series 1200 with a hydrogen flame ionization detector; it was connected to a Speedomax H recorder (1 mv, 1 sec full scale response). The following conditions were used throughout the study;
Injector temperature; 150 C;
Detector temperature; 220 C;
Nitrogen flow rate; 25 ml/min.
In preliminary experiments, packed columns with the following liquid phases were used; 1,2, 3-tris (2- cyanoethoxy)- propane (TRIS);
Carbowax 20M; SF 96-50; OV-17; and diethyleneglycolsuccinate. The
TRIS column gave the best separation of lower-boiling compounds present in strawberry filtrates, and the Carbowax 20M column gave the best separation of the higher-boiling compounds. These two columns were used for comparison of the volatiles of different samples used in this study. Before each use the TRIS column was steam cleaned at 60 C with three on-column injections of 10 |j.l of distilled water at 10 min intervals. The column was then conditioned 1 hr at 28
60 C with a nitrogen flow rate of 25 ml/min. Likewise, the Carbowax column was steam cleaned at 200 C with three on-column injections of
10 JJLI of water at 10 min intervals, followed by conditioning 1 hr at
200 C with a flow rate of 25 ml/min of nitrogen. Column temperature conditions for analysis of volatiles were:
TRIS: 60OC for 48 min;
Carbowax: 50 C for 8 min, then
50-200OC at 40C/min.
Quantitation of GLC Data
Peak areas were calculated by triangulation. The peak height and peak width for shouldering peaks were calculated according to the procedure of Kingston (1964). Peaks appearing before peak 6 on
Carbowax chromatograms were poorly resolved and were not quanti- tated. A computer program (Fortran IV) was written which calculated the peak area and the normalized peak area for each peak. Normalized peak areas were calculated on the basis of internal standard peak area and filtrate yield. Normalized peak areas obtained by use of the program were the average of duplicate runs for each sample. The program also calculated relative retention time (retention time for the peak in question/retention time for the internal standard) as well as total peak area and total normalized peak area for each sample. A detailed description of the program and instructions for its use are included in Appendix I. 29
Statistical Procedures
Analyses of variance were used to compare each peak on the 2 basis of its normalized peak area (in mm ) for different GLC runs.
Not every mold-level classification was processed by every processing
method. Therefore, two different complete sets (three mold levels
and three processing methods, and two mold levels and titiree processing
methods) were analyzed. Since processing and mold-level interactions
were not significant (P > 0.05) in these analyses, mold-level means
were pooled and were compared on the basis of least significant dif- ference (LSD) tests using a pooled estimate of variance from the two
groupings. (See Appendix III.) The effect of processing methods
on the normalized area of each peak was compared in an analysis
of only one mold level (NGQ), and the effects of different maturities were compared in a separate analysis of variance. Mean normalized
peak areas for different processing methods and for different maturi- ties were compared by LSD tests.
Identification of Volatiles
Porapak Collection
In order to increase the amount of the volatiles entrained on the
GLC column and to decrease the amount of water eluting from the GLC 30 column during the GLC-MS run, a modification was developed for the technique of collecting volatiles on a short column packed with a support on which organic compounds have a relatively long retention time and on which water has a relatively short retention time
(Dravnieks and O'Donnell, 1970; Jennings, 1970; Morgan et al. , 1971).
Porapak columns were prepared by filling 4 in x 1 /4 in o. d. stainless steel tubing with acetone-washed 100-120 mesh Porapak Q
(Waters Associates, Inc. , Framingham, Mass. ). The packed
Porapak columns were initially conditioned by placing them in a 200 C oven overnight; when reused, the Porapak columns were conditioned overnight at 150 C. The ends of each Porapak column were labeled so that each column had a no. 1 end and a no. 2 end.
The 20 ml Kimble vial was replaced by a 250 ml reagent bottle containing 125 ml of filtrate.
The procedure for Porapak collection of volatiles was divided into three parts: entrainment of volatiles on the Porapak column, removal of water from the Porapak column, and transfer of volatiles from the Porapak column to the GLC column.
During entrainment of volatiles (Figure 1) the no. 1 end of the
Porapak column was connected to the headspace bottle. The needle through which gases left the headspace vial was connected to the
Porapak column by a 4 in x 1 /8 in o. d. piece of stainless steel tubing heated (by two heat guns) to the same temperature as the water bath. 31
£!RAPA«
ENTRAINMENT GAS IN
0 2 4 i 1 1 1 1 CM
SAMPLE
MAGNETIC STIRRER BAR
Figure 1. Headspace assembly for the entrainment o£ volatiles on the Porapak column. 32
The following conditions were used:
Water bath temperature: TRIS 60 C;
Carbowax 65 C;
Purging gas flow rate: 40 mlAnin of nitrogen;
Entrainment time: 60 min.
Removal of water from the Porapak column with its entrained volatiles was done with the no. 1 end of the Porapak column connected to a nitrogen tank. The column was held at room temperature for
30 min, with a flow rate of 10 ml /min for the first 10 min and a flow rate of 30 ml/min for the last 20 min. The column was then heated to 50 C with a heat gun and was subjected to a flow rate of 25 ml/min.
In transferring the volatiles from the Porapak column to the
GLC column the no. 2 end of the Porapak column was connected to the nitrogen tank and the no. 1 end was connected to the GLC column, the first 10 in of which was U-shaped to accommodate a cold bath (dry- ice and methyl cellosolve). The following conditions were used:
Nitrogen flow rate: 25 ml/min;
Porapak column temperature: 170 C;
Time of collection: 60 min.
The heat was applied to the Porapak column by a rheostat-controlled heat gun. 33
Mass Spectrometry
The instrument, an Atlas CH-3, was a single-focusing mass spectrometer. Volatiles were separated by GLC prior to their entry into the ion source of the MS. The GLC was fitted with a 9:1 splitter so that 10% of the column effluent went to the hydrogen flame detector.
A molecular separator (Llwellyn single-stage silicon rubber mem- brane separator, Varian V-5620, Varian Analytical Division, Palo
Alto, Calif. ) was used at the GLC-MS interface to enrich the GLC effluent in organics by excluding carrier gas. The ionization provided by the 70 eV source was used to obtain mass spectra, which were recorded on a Honeywell 1508 Visicorder. The following operating conditions were used:
GLC
Instrument: F & M 810;
Detector temperature; 260OC;
Injector temperature: 2190C;
Columns:. TRIS; Carbowax;
Column temperatures: TRIS: Isothermal at 60OC for 48 min; Carbowax: Isothermal at 50OC for 8 min; 50OC- 200OCat4OC/min;
Flow rate: 25 ml/min of helium. 34
MS
Filament current: 20 jiA;
Electron voltage: 70 eV;
Accelerating voltage: 3 kV; -7 Analyzer pressure: 5-1/2 x 10 Torr;
Multiplier voltage: 1. 60 kV;
Molecular separator TRIS column: 1780C; temperature: Carbowax column: 186 C;
Scanning speed: 2. 5 sec from m/e 25 to 250, and 5. 0 sec from rn/e. ^ to 250 for compounds with longer retention times.
GLC Retention Times
Retention data for the TRIS colum were used relative to 2- butanone, which was assigned a value of 1. 00. In the same manner retention data for the Carbowax column were used relative to n- heptanol. 35
RESULTS AND DISCUSSION
The analytical procedure using on-column trapping in conjunc- tion with GLC analysis proved effective for comparing the volatiles of different strawberry samples. The headspace on-column entrain- ment method was well suited for obtaining data from many samples, since it was rapid and convenient and required only a small sample.
In contrast, conventional methods using distillation and/or solvent extraction would have required large samples, would have been very time consuming, and would have used elevated temperatures for long periods of time. Conditions such as these could have produced artifact formation. Use of internal standards in the headspace vials allowed compensation for the efficiency of entrainment of different runs, enabling peak areas to be compared.
Identification of compounds was based on comparison of mass spectra with tabulated standard spectra and on correspondence of
GLC relative retention times with those of known compounds. A GLC confirmation (+) indicates that relative retention times for authentic compounds were within 5% of the values for the strawberry compounds.
Tentative MS identifications were given where spectra were weak or were a mixture of more than one compound so that a positive match with reference spectra was not possible. 36
Effect of Mold Level
Mold Counts
The mold count results for berries of different mold levels processed in different ways are shown in Table 1. The subjective mold-level classifications spanned the range used in regulatory work.
For a 50% confidence level a strawberry sample would be accepted for commercial sale at 2 1 or fewer positive fields out of 50 and rejected at 30 or greater (U.S. Department of Agriculture, 1969). If the first counts yielded 22 to 29 positive fields, the sample would be reexamined at a higher confidence level. In this case NGQ berries were acceptable, M and VM berries would have been rejected, and
FQM berries were near the questionable region. Processing method accounted for little difference in mold count for these samples.
Comparison of Mold Cultures
Fifteen different cultures were prepared from VM strawberries and were compared to the A. T. C. C. culture. All strawberry cultures appeared to be wholly populated by B. cinerea, based on microscopic examination. Saboraud maltose broth should not have favored one mold's growth strongly over that of another; it was recommended by
Pelczar (1957) as a general maintenance medium for mold cultures.
Statistical comparison of volatiles after growth of these two cultures 37
Table 1. Howard mold counts (+ fields/50) for strawberries of different subjective mold-level classifications pro- cessed in different ways.
Howard mold count Process NGQ FQM M VM
Fresh 2.3 50.0
Individually Quick Frozen (IQF) 1.5 40.5 50.0 50.0
Frozen Sugared Sliced (FSS) 2. 3 50.0
Freeze Dried (FD) 0.;0u 35.0 50.0
Canned 1.6 32.0 50.0
Preserves 3.0 50.0 38 showed no differences. These results confirmed the work of Powelson
(I960), who found _B. cinerea to be the major mold of strawberries in the Pacific Northwest.
Comparison of Volatiles
TRIS and Carbowax chromatograms comparing the volatiles of
VM and NGQ fruit are shown in Figures 2 and 3, respectively. The normalized peak areas of TRIS and Carbowax peaks for different mold levels are presented in Tables 2 and 3. The identities of these peaks are given in Table 4.
Of major interest is the finding that all volatiles varying signifi- cantly with mold level decreased as mold count increased. No new peaks were evident, and no peaks showed significant increases in level with increased mold growth. These findings cast doubt on the applica- bility of such a HSGLC system for regulatory use. There would be in- herent difficulties in applying a decrease of volatiles as a standard for use in legal proceedings or for routine assay (Campbell, 1970).
The compounds which showed significant (P < 0.05) changes be- tween the different mold-level classifications were n-propionic acid, n-butyric acid, acetaldehyde, isobutyraldehyde, methyl acetate, 1,1- diethoxymethane, 1, 1-diethoxyethane, acetophenone, propiophenone,
1-phenyl-1, 2-propanedione, diethyltoluene, and one unknown. Most of the significant decreases in compounds occurred between NGQ and
FQM fruit; some compounds decreased between FQM and M fruit, but II 12 13
20 25 30 35 40 45 50 TIME (MIN)
8 9 10 II 12 13 15 I5A x32 x2 I0A ft
x?fifi 3 x2 " M ft A I2A I" \l IAI
20 25 30 35 40 45 50 TIME (MIN) Figure 2. Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a TRIS column; top, very moldy (VM) fruit, and bottom, normal good quality (NGQ) fruit. 26A IBBI8D 20A X32 6A ISA100 19 20 21 22 27 34 36 39A x4 X8x4x4)i4 x2 I x2 x2 ft UJ (/) \ z 5 x64 o XI0Z4 A A Q o: o o UJ IT 20 25 30 35 40 45 50 TIME (MIN) 180 9 10 1112 16 I8AI8B 19 20 22 x2 xl6 x2 x4)i4 x4 UJ CO oZ UJ Q Q:o o a:UJ If W 20 25 30 35 40 45 50 TIME (MIN) Figure 3. Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a Carbowax column; top, very moldy (VM) fruit, and bottom, normal good quality (NGQ) fruit. o H R re p - •V 3 a oo s vj in *. w VO 00 00 C» SI v) CT\ o^ in w to i-' O O O B (0 o A A > ^ w > > > w > cr sr *-* O ST 5 ~s re re O to o in •z re M o M *. Mi-'l-'l-'l-'tOUlVOl-'Ml-'tvJtOWlX)*. 00 4». !-» p i^. -J 00 Ul »-» Oi *. lU >-» Vl O JO 00 _N> N VO N> O H* *>. O lu lu in Z re l-»(O00v)VOI-'vJ(« ^OO^tOtOvJ^OO W (O I-' W KD 00 IU o vi in ■t^ooooNtoooinvoO 6 W in \0 Ol if>- 00 CTl VI in vo s re s a 13re s p 13 3 jxr 1-^ I-* IO t-^- 00 (\} H* W N H^ ^ ^ o O il^ 00 o\ *. O 00 P VO to ►-» o ►-» Ol-'tOUJOONlvJ p s ■O O in Oi ooiDOi-'vji-k!-»<*>ooC)VOOlv0v0l(^l0v0t_k,_itoin.t^intoooo>Oi O v) oo ^. O V] in IW tO VO PJ., V] ^l-itOvltOWOOtOtOOllUtOlMvllMil^OOl-i s ep £ re o a to H to to 13 >-»to V)I-»I-K toto win o> re 2 oi oo o vo winv)v]y5in00IOUJvjvjOH»i-*v) - 6 cuOOi-'U3^u>^. p 00 sr «? vo in Oi OJ vointoOiwinvoit^CM ^►-'VOviinvoooOoo i-» o^ ib. vi K* O vo to O vi O to vlUJtOi-'v)^i-iv)0 tooituvoint-'ootnto H* O 0\ O *. in a. pa. to tO O P I i-'toi-'00i-»i-'i-»H»to I-^OJ in in *> i-» B cr oo oo o i-' inv]00iotioOvo>l^(/iOiiu»-*h-'Vo OOi-'i-»t0OO*.O w OJ in a. o ft *.i-'00inOUJv)Ovoi-»*.ooO to i-» oo oi OJ vo O to vo in ■b. IU *. oo. v] l-» o vjtomwvoOi-'Ovoooinoovi ft vo vi o t-» o to vo oo in A oo £ vo to o 3 •e ga P n in O v) Ol into>-' b.ioovov) ^oioiotooo O v] O O Ol VO O i-k to i-» o o l l sI IM I-' ■b. >-» vo O t— in ibOO100Oll-^0000l->OV0 Table 3. Normalized peak areas of Carbowax volatiles and comparison of different mold levels. Mean normalized peak area (cm ) Comparison of means ; Peak number NG 6 647. 62 580. 77 511.23 492. 73 -66. 85 -69. 54 -18. 50 6A 337. 20 352.12 326. 45 331.12 14.92 -25.67 4.67 9 42.91 30.96 17.57 14.10 -11.95* -13.39* - 3.47 10 182. 58 106.78 42.31 12.79 -75. 80* -64. 47* -29. 52 11 135. 86 114.25 87.46 69.38 -21.61 -26. 79 -18.08 12 7.32 3.70 1.31 1.15 - 3.62* - 2.39* - 0.16 14 6.75 5.02 5.16 4.87 - 1.73 0.14 - 0.29 16 29.05 16.49 8.09 7.50 -12/56** - 8.40* - 0.59 18 4.27 3,75 3.81 3.29 - 0.52 0.06 - 0.52 18A 10.97 12.09 11.17 9.45 1.12 - 0.92 - 1.72 18B 19.14 17.43 17.06 13.87 - 1.71 - 0.37 - 3.19 18C 6.46 5.96 5.52 4.93 - 0.50 - 0.44 - 0.59 18D 33.86 27.92 23.14 18.56 - 5.94 - 4.78 - 4.58 19 17.03 17.54 15.47 11.75 0.51 - 2.07 - 3.72 20 28. 49a 27.61a 27. 93a 29. 52a - 0.88 0.32 1.59 21 6.38 6.78 5.88 4.61 0.40 - 0.90 - 1.27 22 42.91 35.27 30.16 33.14 ^ 0.64 - 5.11 2.98 22A 7.81 6.94 6.50 6.08 - 0.87 - 0.44 - 0.42 23 9.38 7.90 6.81 7.25 - 2.48 - 1.09 0.44 25 13.54 12. 78 13.32 10.42 - 0.76 0.54 - 2.90 26 12.80 9.14 10.66 8.90 - 3.66 1.52 - 1.76 26A 8.06 6.89 6.01 6.57 - 1.17 - 0.88 0.56 27 9.48 10.13 8.73 9.60 0.65 - 1.40 0.87 28 8.42 8.82 8.51 6.15 0.40 - 0.31 - 2.36 29A 2.76 2.65 2.45 2.05 - 0.11 - 0.20 - 0.40 30 1.95 2.06 1.80 2.23 0.11 - 0.26 0.43 32 10.28 9.81 10.57 9.84 - 0.47 0.76 - 0.73 34 26.46 27.63 23.42 18.11 1.17 - 4.21 - 5.31 35 5.73 5.60 6.17 6.65 - 0.13 0.57 - 0.48 36 110.19 95.85 99.30 87.15 -14.34 3.45 -12. 15 37 8.94 10.68 9.72 11.47 1.74 - 0.96 1.75 39A 162.96 189.24 165. 43 146.18 26.28 -23.81 -19.25 Peak: area represents 0. 1 ppm of internal standard (n-heptanol). *P <0. 05 **P <0. 01 Table 4. GLC-MS identification of volatiles important in mold-level differences. Peak MS MS GLC Compound number identification reference confirmation TRIS 2 n-Propionic acid + A. S. T.M. , 1969 + 4 n-Butyric acid + ibid. + 6 Acetaldehyde ■ + ibid. + 7A Isobutyraldehyde + ibid. + 8 Methyl acetate + ibid. + 10A 1,1- Diethoxymethane + McFadden et al. , 1964 + 12A Unknown 15A 1,1- Diethoxyethane + ibid. + Carbowax 9 Acetophenone + A. S. T.M. , 1969 + 10 Propiophenone + ibid. + 12 1-Phenyl- 1, 2-propanedione + ibid. 16 Diethyltoluene tentative ibid. 43 none decreased significantly from M to VM fruit. Compounds de- creasing significantly from NGQ fruit to FQM fruit -were: n-propionic acid, n-butyric acid, acetaldehyde, isobutyraldehyde, methyl acetate, 1, 1-diethoxymethane, 1, 1-diethoxyethane, acetophenone, propio- phenone, 1- phenyl- 1, 2-propanedione, and diethyltoluene. Those compounds decreasing significantly from FQM to M fruit -were as follows: acetaldehyde, methyl acetate, 1, 1-diethoxymethane, 1,1- diethoxyethane, acetophenone, propiophenone, 1 - phenyl- 1, 2-propane- dione, and dithyltoluene. Sandoval and Salwin (1968) found one GLC peak characteristic of decomposed strawberries which persisted after FD. However, this peak was not identified and could have formed during storage for 10 days at 10 C rather than being caused by mold growth itself. Salwin (1969) found that volatile reducing substances of decomposed peaches, shrimp, and ground beef were partly lost when the foods were cooked and were completely lost when the foods were FD. The decrease in the level of butyric and propionic acids on mold growth agrees with similar findings for B. cinerea and other micro- organisms. Minarik (196l) found that growth of _B. cinerea decreased the level of volatile acids in grapes. Gray (1959) and Amerine and Cruess (I960) reported the same results. Charpentie (1954) found that B. cinerea oxidized volatile acids, particularly at the low pH's found in fruits. 44 Niewiadomski and Salmonowicz (1969) found Penicillium and Fusarium used volatile acids. Reddy ^t al. (1968) found that Pseudomonas fragi converted butyric acid to ethyl butyrate. The decrease in acetaldehyde and isobutyraldehyde with increas- ing mold count is substantiated by previous findings on the use of aldehydes by microorganisms. Hansen and Keeney (1970) found that saturated aldehyde concentrations were lower in moldy than in non- moldy cocoa beans. Fields ejt al. (1968) found that acetaldehyde was used in the synthesis of acetylmethylcarbinol by molds. Stevens et al. (1969) found that aldehydes were lost in the fermentation of wines by yeast. Keenan and Bills (1968) found that five species of Pseudomonas had the capacity to reduce acetaldehyde and butyraldehyde to their corresponding alcohols. The loss of methyl acetate associated with B. cinerea growth could have been caused by a mold-produced esterase. Pitt (1968) reported that B. cinerea produced an esterase; Reddy et al. (1968) found that Pseudomonas fragi produced intracellular esterases. A decrease in the concentration of the acetals 1, 1- diethoxymethane and 1, 1-diethoxyethane was associated with B. cinerea growth. Popova and Puchkova (1953) found that the growth of B. cinerea on champagne grapes decreased the concentrations of lower molecular weight acetals. 45 Little work has been done on the metabolic breakdown of aromatic compounds such as acetophenone, propiophenone, and 1-phenyl- 1, 2- propanedione by molds such as Botrytis. Henderson and Farmer (1955) found that a Botrytis sp. metabolized aromatic carbonyls by conversion to aromatic acids and then by breaking the benzene ring of the aromatic acid. The reaction was not further elucidated. Gross _et al. (1968) found that Neurospora crassa contained two distinct enzymes, an aryl-aldehyde: NADP oxidoreductase and an aryl-alcohol reductase, which together were capable of reducing aromatic acids to the corresponding alcohols. Either type of reaction could have result- ed in the reduction of the level of propiophenone and 1-phenyl-1, 2- propanedione associated with B. cinerea growth. Nambudiri et al. (1970) found that an Alternaria sp. utilized several aromatic alde- hydes and aromatic acids and proposed that their use functioned as a detoxification mechanism for the mold. A decrease in the level of diethyltoluene was associated with an increase in mold level. A limited ability of molds to use aromatic hydrocarbons was found by Tanaka et al. (1968) and was thought to function as a detoxification mechanism; such a usage may have resulted in the reduction in level of diethyltoluene. Table 5 shows the TRIS compounds whose decrease was associated with mold growth (Group A) and the compounds to which they could have been altered by mold action (Group B). The levels of Table 5. Comparison of relative amounts of TRIS compounds decreasing on mold growth with relative amounts of the products to which they could have been altered. Group A Group B Compounds decreasing with mold growth Possible products of mold act ion TRIS MS % of TRIS MS % of peak no. identification TNPA peak no. identification TNPA 4 Butyric acid 0.2 16 Ethyl butyratec 5. 3 6 Acetaldehyde 0.6 10 Ethanold 32.8 7A Isobutyraldehyde 0. 5 12A Isobutyl alcohol 15.8 8 Methyl acetate 2. 1 Methanol6 1 Acetic acid6 0.2 10A 1, 1-Diethoxym ethane 0.2 10 Ethanolf 32.8 Formaldehyde 15A 1, 1-Diethoxyethane 0. 5 10 Ethanolf 32.8 6 Acetaldehyde 0. 6 Amounts of both Group A and Group B compounds are expressed as % of total normalized peak area in fresh strawberries. Total normalized peak area :Reddyetal. (1968) Keenan and Bills (1968) JPitt (1968) f Popova and Puchkova (1953) ►^ o^ 47 compounds in both groups in fresh strawberries are also given in the table. These results could have accounted for failure to find signifi- cant increases in any Group B compounds associated with mold growth. The Group B compounds were usually present in strawberries in much larger amounts (5 to 30% of total normalized peak area, TNPA) than the group A compounds (0. 2 to 2. 1% of TNPA). Thus, stoichiometric formation of Group B compounds from Group A compounds would have resulted in relatively small increases in the level of Group B compounds compared to the amounts of Group B compounds already present in strawberries. Mold growth was found to be associated with a significant (P < 0.05) decrease in the level of methyl acetate (2. 1% of TNPA), but not in the level of ethyl acetate (26.8% of TNPA). Possibly there was insufficient mold esterase to cause a significant (P < 0.05) decrease in the level of ethyl acetate during the period of mold growth. It is also possible that an esterase with different affinities for the two esters could have accounted for this finding. It is important that ethanol and diacetyl were herein identified by MS and by GLC retention times (TRIS peaks 10 and 16, respectively) and that significant increases in their concentrations were not associated with increasing mold counts. The results of this study thus substan- tiate the finding by Fields et ah (1968) that ethanol and diacetyl were not indicators of B. cinerea spoilage. They proposed that an increase 48 in the level of ethanol or diacetyl could be used as an indicator of some types of microbial spoilage of foods, but they did not find B. cinerea to be a producer of either compound. The study here under- taken indicates that the complexity of biological material may preclude the possibility of finding a single indicator of mold decomposition and that changes in several compounds may have to be considered. Patterson (1945) suggested that the delay in accepting chemical indexes of microbial spoilage has been caused by the desire to find one compound which could be used universally. Of the compounds whose alteration in level was found to be associated with mold growth, the following have been approved for food use: acetaldehyde, isobutyraldehyde, butyric acid, and methyl acetate. All of these have been approved for use as synthetic flavor- ing agents (Furia, 1968). Fields ^t al. (1968) advised that compounds used in food processing should not be used as indicators of microbial quality. Effect of Processing Method Figure 4 compares TRIS chromatograms for freeze dried and fresh fruit; similarly, Figure 5 gives TRIS chromatograms for preserves and fresh fruit. Carbowax chromatograms for freeze dried versus fresh fruit (Figure 6) and for canned versus fresh fruit (Figure 7) are also provided. The TRIS peaks which varied significantly 2 10 12 »4 UJ V) 2 O Q. (/) UJ tr ' tr UJ o tr o o UJ l»A'OB tr OA ^j 8A8B 10 15 20 25 30 35 40 45 50 TIME (MIN) lO I 2 8 9 10 II 12 13 15 ISA |x2 x2xl6 x32 III UJ (/) 2 O a. to , 0B tr 0A 7A ui Q o o 6A UI A l 7 8A XL w 10 20 25 30 35 40 45 50 TIME (MIN) Figure 4. Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a TRIS column; top, freeze dried (FD) fruit, and bottom, fresh fruit. 25 30 50 TIME (MIN) lO I 2 6 8 9 10 II 12 13 15 ISA «2 x2xl6 x32 x2 x8 IOA rt n n x?56 fM x2 ft n x4 ' I2A i V l> U IAI 14 /BB 8A 10 15 20 25 30 35 40 45 50 TIME (MIN) Figure 5. Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a TRIS column; top, preserves, and bottom, fresh fruit. UJ (A) Z o Q. C/> UJ ST. (T UJ Q tr o o UJ tr 20 25 30 TIME (MIN) I8D 12 6 6A 9 10 II 12 16 I8AI8B 19 20 22 «256 XI024 x2 xl6 x2 x4x4 x4 UJ %2 zCO o »256 1(128 cn )«32 A UJ Offl xl2E tr UJ Q 10 20 25 30 35 40 45 50 TIME (MIN) Figure 6. Analysis of the headspace volatiles present in an aqueous extract of normal good quality (NGQ) strawberries using a Carbowax column; top, freeze dried (FD) fruit, and bottom, fresh fruit. RECORDER RESPONSE RECORDER RESPONSE X — 3 -8-"1 °"k i ro> a) PO •> "i- w> ro (TO w Ol 5~ UI9 ^ e "^K- 0) -ir— B ►a 3 c ct P^. 0 0 3 •p CO " a o o (0 a 0 i-^ »-»! 4 c 3 0 rt- 3*-t 3 pj a era cr o 0 0 rt- a. 0 25 53 (P < 0. 05) with processing method are listed in Table 6; each pro- cessing method is compared to fresh NGQ strawberries. A similar summary for Carbowax peaks is presented in Table (. Tables 8 and 9 contain the identities of these TRIS and Carbowax peaks respectively. Effect of Freezing (IQF and FSS) The loss of volatiles due to these two freezing processes was confined to the lower boiling TRIS volatiles (Tables 5 and 6). There was a significant (P < 0. 05) decrease in five peaks for the IQF process and in two peaks for the FSS process. Compounds which decreased significantly as a result of the IQF process were as follows: diethyl ether (TRIS peak 0); acetic acid (TRIS peak 1); propionic acid (TRIS peak 2); n-butyric acid (TRIS peak 4); and acetaldehyde (TRIS peaK 6). Diethyl ether (TRIS peak 0) and acetic acid (TRIS peak 1) showed significant decreases due to the FSS process. Since the IQF fruit was directly exposed to an air blast overnight -while the FSS fruit was protected by the fiber container, the greater loss of volatiles occurring with IQF fruit seems reasonable. Additionally, some loss of volatiles could have occurred during the two months of frozen storage prior to HSGL.C analysis. The sugar sirup used in the FSS fruit could also have been a protecting factor (Ponting et al. , 1968). Table 6. Nonnalued peak areas of TRIS volatiles and comparison of piocessed with fresh fruit. 2 Mean normalized peak areas (cm ) Comparison of means Peak Process level of number Fresh I(?F FSS FD Canned Preserves IQF-Fresh FSS-Fie* FD-Fresh Canned-Fresh Preserves-Fresh significance 0 9.84 5.16 5.60 2.72 5.24 4.96 - 4.68* - 4.24* _ 7.12** - 4.60* _ 4.88* ** QA 4.02 3.42 3.70 3.65 4.18 3.55 - 0.60 - 0.32 - 0.37 0.16 - 0.47 OB 5.14 4.76 4.92 4.61 4.87 4.58 - 0.38 - 0.22 - 0.53 - 0.27 - 0.S6 1 16.20 10.60 11.41 2.58 10.74 8.87 - 5.60* - 4.79* - 13.62** - 5.46* - 7.33* ** 2 158.24 92.16 140.22 18.29 86.32 70.64 -66.08** -18.02 - 139. 95** - 71.92** - 87.60** ** 3 0.48 0.45 0.44 0.44 0.51 0.46 - 0.03 - 0.04 - 0.04 0.03 - 0.02 4 3.16 1.50 2.89 0.82 1.58 1.34 - 1.66* - 0.27 - 2.34** - 1.58* - 1.82* ** 5 0.50 0.72 0.55 0.48 8.94 6.46 0.22 0.05 - 0.02 8.44** 5.96** ** 6 218.69 129.60 205.87 110.56 162.81 140.02 -90. 75** -12.82 - 108.13** - 55.88** - 78. 67** ** 6A 1.22 0.92 1.16 1.30 1.18 1.34 - 0.30 - 0.06 0.08 - 0.04 0.12 7 0.91 0.80 0.88 0.94 1.04 1.12 - 0.11 - 0.03 0.03 0.13 0.21 7A 2.60 2.34 2.51 2.45 2.46 1.54 - 0.26 - 0.11 - 0.15 - 0.14 - 1.06* * 8 269.42 284.48 286.04 31.22 137.58 127.65 15.06 16.62 - 238.20** -131.84** - 141. 77** ** 8A 2.14 1.98 2.02 1.87 2.05 1.95 - 0.16 - 0.12 - 0.27 - 0.09 - 0.19 8B 2.20 1.95 2.31 1.92 2.11 2. OS - 0.25 0.11 - 0.28 - 0.09 - 0.15 9 3590.10 3502.08 3624.81 58.36 3395.10 520.27 -88.02 34.71 -3531. 74** -195.00 -3069. 83** ** 10 480.02 459.96 475. 78 143.70 102.58 94.66 -20.06 - 4.24 - 336. 32** -377.44** - 385. 36** ** 10A 27.13 23.63 28.02 1.89 24.76 15.87 - 3.50 0.89 - 25.24** - 2.37 - 11.26** ** 11 41.20 38.48 37.45 4.16 39.51 2.53 - 2.72 - 3.75 - 37.04** - 1.69 - 38.67** ** 12 20.40a 21.06a 22. 84a 23.27a 22.123 19.21* 0.66 2.44 2.87 1.72 - 1.19 12A 14.03 11.88 13.72 12.38 15.17 16.10 - 2.15 - 0.41 - 1.65 1.14 2.07 13 25.17 28.56 26.19 16.70 23.96 24.28 3.39 1.02 - 8.47* - 1.21 - 0.89 * 14 18.12 16.40 19.78 10.32 17.49 15.88 - 1.72 1.66 - 7.80* - 0.63 - 2.24 * 15 125. 88 117.60 130.52 29.44 114.22 32.75 - 8.28 4.64 - 96.44** - If. 66 - 93.13** ** ISA 49.51 44.16 47.75 18.65 45.87 5.22 - 5.35 - 1.76 - 30.86** - 3.64 - 44.29** ** 16 35.42 30.90 31.00 22.08 33.80 20.64 - 4.S2 - 4.42 - 13.34* - 1.62 - 14.78* * 17 19.65 18.43 21.72 9.47 19.89 17.48 - 1.22 2.07 - 10.18* 0.24 - 2.17 * 17A 17.09 14.25 14.04 11.38 16.42 15.65 - 2.84 - 3.05 - 5.71* - 0.67 - 1.44 * 18 12.48 11.02 13.43 4.27 13.81 10.92 - 1.46 0.95 - 8.21* 1.33 - 1.56 * 18A 13.02 11.25 13; 50 4.59 11.96 12.49 - 1.77 0.48 - 8.43** - 1.06 - 0.S3 ** 19 29.68 29.40 27.52 13.16 26.04 22.16 - 0.28 - 2.16 - 16.52** - 3.64 - 7. 52* ** Peak area represents 0. 5 ppm of internal standard (2-butanone). *P<0.0S **P<0.01 Table 7. Normalized peak areas of CaAowax volatiles and comparison of processed with fresh fruit. 2, Mean normalized peak areas (icm ) Comparison of means Peak Process level of number Fresh IQF FSS FD Canned Preserves IQF-Fresh FSS-Fresh FD-Fresh Canned-Fresh Preserves-Fresh significance 6 894. 32 860.16 842. 55 192.75 820.25 275. 67 -34.16 -51.77 -701. 57** -64.07 -618. 65** ** 6A 136.47 132. 56 141. 72 42.37 856.22 714. 25 - 4.31 5.25 - 94.10* 719.75** 577. 78 ** 9 51.76 48.96 53.28 24.61 46.73 32.14 - 2.80 1.52 - 27.15** - 5.03 - 19.62** ** 10 217.45 221. 76 234.16 71.22 198.19 152.67 4.31 16.71 -146.23** -415.64** - 64. 78* ** 11 222. 81 217. 60 228. 87 42.88 24.28 78.70 - 5.21 6.06 -179. 93** -198. 53** -144.11** ** 12 11.47 9.36 10.14 3.25 5.42 4.25 - 1.33 - 1.33 - 8.22** - 6.05* - 7.22* ** 14 8.02 6.95 7.46 4.30 10.71 3. OS - 1.07 - 0.56 - 3.72* 2.69 - 4.97* * 16 39.19 41.28 35.77 8.15 33.12 16.76 2.09 - 3.42 - 31.04** - 6.07 - 22. 43** ** 18 5.20 4.85 5.65 2.49 4.61 2.80 - 0.35 0.45 - 2.71* - 0.59 - 2.40* * ISA 12.61 12.90 10. 76 7.21 14.89 7.46 0.29 - 1.85 - 5.40* 2.28 - 5. IS* * 18B 25.12 22.80 24.14 6.17 24.13 12.49 - 2.32 - 0.98 - 18.95** - 0.99 - 12.63* ** 18C 8.06 6.45 7.20 5.23 6.81 4.98 - 1.61 - 0.86 - 2.83* - 1.25 3.08* * 18D 33.88 37.24 30.17 1.98 48.17 51.73 3.36 - 3.71 - 31.90** 14. 29* 17 85* ** 19 21.85 19.60 19.98 5.15 26.35 9.27 - 2.25 - 1.87 - 16.70** 5.50 - 12.S8* ** 19A 0.00 0.00 0.00 0.00 6.83 8.22 0.00 0.00 0.00 6. 83** 8.22** ** a a a 20 27.05* 28.60* 24.22 30s47 31. 14* 29. 48 1.55 - 2.83 3.42 4.09 2.43 21 6.71 5.38 5.74 6.32 8.73 5.42 - 1.33 - 0.97 - 0.39 2.02 - 1.29 22 43.06 39.36 40.93 21.35 38.26 37.25 - 3.70 - 2.13 - 21. 71** - 4.80 - S.81 ** 22A 8.60 7.68 8.05 7.93 7.45 7.14 - 0.92 - 0.55 - 0.67 - 1.15 - 1.46 23 11.26 10.32 12.37 2.52 9.94 9.86 - 0.94 1.11 - 8.74** - 1.32 - 1.40 ** 25 15.14 13.80 14.50 12.81 12.32 12.65 - 1.34 - 0.64 - 2.33 -2.82 - 2.49 26 13.02 , 10.45 12.46 10.16 16.49 14.20 - 2.57 - 0.56 - 2.86 3.47 1.18 26A 8.43 8.72 7.92 6.31 8.02 8.93 0.29 - 0.51 - 2.12 0.50 0.50 27 8.99 10.62 8. 5S 9.47 10.98 8.27 1.63 - 0.44 0.48 1.99 - 0.72 28 6.14 5.45 5.88 6.02 14.68 12.05 - 0.69 - 0.26 - 0.12 8.54** S.91* ** 29A 2.90 2.38 2.46 2.71 3.52 2.59 - 0.52 - 0.44 - 0.19 0.62 - 0.31 30 1.96 2.17 1.82 2.02 2.08 1.63 0.21 - 0.14 0.06 0.12 - 0.33 32 9.43 10.60 11.18 10.94 11.24 8.27 1.17 1.75 1.51 - 1.16 - 1.16 34 25.20 21.70 22.77 29.47 30.49 29.12 - 3.50 - 2.43 4.27 5.29 3.92 35 6.52 5.40 6.03 5.16 4.81 6.47 - 1.12 - 0.49 - 1.36 - 1.71 - 0.05 36 121.24 110. 88 115.66 98.14 109.40 105. 83 -10.36 - 5.58 - 23. 10 -11.84 - 15.41 37 10.19 8.64 9.47 8.45 8.18 8.72 - 1.55 - 0.72 - 1.74 - 2.01 - 1.47 39A 161.75 146.00 157.23 181.53 178. 45 152.80 - 15. 75 - 4.52 19.78 16.70 - 8.95 Peak area represents 0.1 ppm of internal standard (n-heptanol). *P <0.05 **P <0.01 Table 8. GLC-MS identities of TRIS volatiles differing in amount with processing method. Peak MS MS GLC Compound number identification reference confirmation 0 Diethyl ether + A. S.T.M. , 1969 + 1 Acetic acid + ibid. + 2 Propionic acid + ibid. + 4 n-Butyric acid + ibid. + 5 Dimethyl sulfide + ibid. + 6 Acetaldehyde + ibid. + 7A Isobutyraldehyde + ibid. + 8 Methyl acetate + ibid. + 9 Ethyl acetate + ibid. + 10 Ethanol + ibid. + 10A 1, 1-Diethoxymethane + McFadden et al. , 1964 + 11 Ethyl propionate + A. S.T.M. , 1969 + 13 Methyl butyrate + ibid. + 14 1, 1-Dimethoxyethane + McFadden et al. , 1964 + 15 Ethyl butyrate + A.S.T.M., 1969 + 15A 1, 1-Diethoxyethane + McFadden et al. , 1964 + 16 Diacetyl + A. S.T.M. , 1969 + 17 1-Butanol + ibid. + 17A 1, 1-Ethoxypropoxyethane + McFadden et al. , 1964 18 1, 1-Diethoxypropane + ibid. + 18A 1, 1-Ethoxypentoxyethane + ibid. 19 2-Methyl-2-butanol + A. S.T.M. , 1969 + U1 Table 9. GLC-MS identities of Carbowax volatiles differing in amount with processing method. Peak MS MS GLC Compound number identification reference confirmation 6 Ethyl butyrate + A.S. T. M. 1969 + 6A Benzaldehyde + ibid. + 9 Acetophenone + bid. + 10 Propiophenone + ibid. + 11 1 - Phenylbutan- 1 - one + ibid. 12 1-Phenyl- 1,2-propane dione + ibid. 14 3- Phenylpropan-2- one + ibid. 16 Diethytoluene + ibid. 18 Unknown 18A o-Cymene + ibid. 18B p-Cymene + ibid. 18C Unknown 18D Furfural + ibid. + 19 Linalool + ibid. + 19A 5-(Hydroxymethyl)-2- furfural + ibid, 22 n-Pentyl n-hexanoate + ibid, 23 cis-3-Hexen-1-yl hexanoate + ibid. 28 1- Propanethiol + ibid. 58 Dimick and Makower (1956) reported formation of_n-hexanal from 2-hexenal during frozen storage of strawberries. Workers at the Sprenger Institute (1969) substantiated this finding and proposed that formation of n-hexanal accounted for the off-flavor associated with frozen strawberries. Added sucrose was found to inhibit the forma- tion of this off-flavor (ibid. ). In the present study no significant increase (P < 0.05) was found in any of the peaks with either process. The short interval of frozen storage (two months) before GLC analysis could have accounted for lack of formation of new peaks. Strawberries used by Dimick and Makower (1956) and by the workers at the Sprenger Institute (1969) were stored for one year. A longer period of frozen storage than that used in the present study may have been needed for significant increase in the level of ji-hexanal. Effect of Freeze Drying The FD fruit showed a significant decrease for almost all peaks in the TRIS chromatogram and a significant decrease in TRIS TNPA (Table 5 and Figure 3). With the Carbowax chromatogram there was significant decrease for the lower-boiling half of the Carbowax pattern of volatiles (peaks 6 through 19) and for Carbowax TNPA (Table 6 and Figure 5). These results agree with those of several other studies on the effect of freeze drying on volatiles. Saravacos and Moyer (1968) found 59 a significant loss of flavor compounds on freeze drying in model systems. Heatherbell ^t al. (1971) found that freeze drying caused loss of 75% of the total volatile content of carrots. Strawberry volatiles would be expected to be very susceptible to loss on freeze drying, as many of them have very low vapor pressures (Nursten and Williams, 1967). Effect of Canning Canning caused loss of lower-boiling TRIS volatiles (Table 5) and heat induction of dimethyl sulfide (Table 5) and benzaldehyde, furfural, 5- (hydroxymethyl)-2-furfural, and propanethiol (Table 6 and Figure 6). Total normalized peak area for TRIS compounds significantly decreased on canning. Loss of lower-boiling volatiles could have resulted from heating of strawberries in the unsealed can during the exhaust process. Dimick and Makower (1956) found that heating of strawberry puree resulted in loss of lower-boiling volatiles. Comparable losses have been found during blanching of beans (Matthews, I960), canning of carrots (Heatherbell et ah , 1971), heating of bananas (Hultin and Proctor, 196l), and canning of tomatoes (Katayama et al. , 1967; Nelson and Hoff, 1969; Stevens, 1971). Dimethyl sulfide has been reported as a heat-induced compound in several different food products. In the following cases it has been found in the heated product but has been essentially absent from the 60 raw, fresh product; strawberries (Sloan etal. , 1969); potatoes (Self, Rolley and Joyce, 1963); beans (Self, Casey and Swain, 1963); tomatoes (Guadagni et aL , 1968; Nelson and Hoff, 1969); cabbage and cauliflower (MacLoed and MacLoed, 1970); corn (Bills and Kennan, 1968); and carrots (Heatherbell et al. , 1971). S-methylmethionine sulfonium salt has been found to be the precursor of dimethyl sulfide in canned tomatoes (Wong and Carson, 1966), in heated milk (Keeoan and Lindsay, 1968), and in canned corn (Bills and Keenan, 1968). In this study benzaldehyde increased significantly with canning. It has been reported as a heat-induced compound in milk (Scanlan et al. , 1968) and in cocoa (Marion et al. , 1967). Hodge (1967) identified benzaldehyde in model system studies on the caramelization of sugars; its immediate precursor was not found. Benzaldehyde has also been reported in several cases to have been formed by processes other than heat induction. Jennings and Sevenants (1964) found benzaldehyde in freestone peaches and proposed that it originated in the stone rather than in the flesh of the fruit. Liener (1966) reported that benzaldehyde was formed by the hydrolysis of cyanogenetic glycosides in almonds, stone fruit kernels, and lima beans. This hydrolysis was found to be effected by mildly acidic conditions or by enzymes called "emuj.sins, " which were nearly always contained in plants having cyanogenetic glycosides. In a study con- ducted by Parks and Patton (1961) benzaldehyde was found to have arisen as a result of feed constituents entering milk. 61 Furfural increased in concentration due to canning. The trace amount present in fresh strawberries could have been an artifact caused by the heating in the on-column entrainment process. Furfural has been found as a heat-induced product in strawberries (Sloan et al. , 1969), maple sirup (Filipic et al. , 1969), snapbeans (Stevens et al. , 1967), and milk (Parks, 1967). Hodge (1967) reported furfural to be a nonenzymatic browning product formed by dehydration of pentoses. Xylose and arabinose, subunits of polysaccharides in plant cell walls (Albersheim, 1965), could have been precursors of furfural. Williams et al. (1952) found xylose present as a free sugar in straw- berries to the extent of 0. 1% by weight of the fresh fruit. In this study 5- (hydroxymethyl)-2-furfural was formed in the canning process. Filipic et al. (1969) found that this compound was formed by the heating process in maple sirup. It has been reported as a dehydration product of hexoses (Hodge, 1967), particularly glucose (Walter and Fagerson, 1968). Glucose and fructose are the only hexoses which have been reported as free sugars in strawberries. Widdowson and McCance (1935) found that glucose and fructose represented 2. 59% and 2. 32%, respectively, of the weight of the edible portion of fresh strawberries. Free hexoses other than glucose and fructose were reported to be found rarely in fruits and then only in trace amounts (Whiting, 1970). However, other hexoses such as 62 galactose and mannose have been reported as subunits of cell-wall polysaccharides (Albersheim, 1965). Propanethiol increased upon canning. Similar findings have been reported by Self (1967) in his work with potatoes, beans, and corn. Heatherbell et_ aL (1967) reported formation of ethanethiol in carrots upon canning. No work on the precursor of propanethiol in foods has been reported. Methanethiol is formed on heating of 3-methyl- thiopropanal (Day, 1967). Gumbmann and Burr (1964) proposed that sulfur-containing volatiles in foods originated primarily from the sulfur-containing amino acid residues of proteins, although other substances such as thiamin, biotin, co-enzyme A, and glutathione were also listed as possible precursors. Effect of the Preserving Process This process caused significant loss of most TRIS volatiles (Table 5 and Figure 4) and loss of many Carbowax volatiles (Table 6). Total normalized peak area for TRIS and Carbowax volatiles was decreased significantly by preparation of preserves. These results could have occurred due to the heating (880C maxi- mum) and the vacuum (640 mm) applied. Not as many compounds were lost in preparing preserves as in freeze drying; the loss of volatiles for preserves was slightly greater than for canned fruit. 63 Severe loss of volatiles has been found previously in the processing of strawberry jam (Katayama et al. , 1968). Compounds showing significant increase in the preserving process were dimethyl sulfide (Table 5 and Figure 4) and benzaldehyde, furfural, 5-.(hydroxymethyl) -2-furfural, and propanethiol (Table 6). The same compounds increased significantly upon canning. No other compounds were found to increase significantly upon the manufacture of preserves. Effect of Maturity TRIS and Carbowax chromatograms comparing UR, NGQ, and OR fruit are shown in Figures 8 and 9. The TRIS peaks which varied significantly (P < 0. 05) with maturity are presented in Table 10. A similar summary for Carbowax peaks is shown in Table 11. The identities of these TRIS and Carbowax peaks are given in Tables 12 and 13, respectively. In general most compounds which changed significantly (P < 0.05) increased with maturity. No new compounds were formed with ripen- ing, and no compounds completely disappeared with ripening. This agrees in part with the work of Ahmed and Scott (1963), who studied Fairfax and Dixieland strawberries. They found that the volatiles present in unripe fruit increased with ripening; however, they reported that several new compounds were formed, none of which were identified. 64 9 10 II 12 13 <3 2 n DA " 25 30 TIME (MIN) x2 II «R IDA n n «?5S f A «2 l\ n s4 { ISA n 14 V 25 30 TIME (MIN) 0 I 2 <2 x2;8 UJ(/) oZ CL 10 Q tr o U1 20 25 30 TIME (MIN) Figure 8. Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a TRIS column; top, underripe (UR) fruit; middle, normal good quality (NGQ) fruit, and bottom, overripe (OR) fruit. 65 I8B 180 10 II 16 ISA ISC 19 20 x2 x2 x4 s4 x4 x4 OA COz o P , CL 3 tn .512 xl6 A 1 2 I F Q IT O o LLJ IT 20 25 30 TIME (MIN) 12 6 6A 9 10 II 12 16 ISA IBB 19 20 22 <2S6 XI024 x2 xl6 x2 x4rf x4 2 z ■ o b Q- (/> x2S6 UJ <32 « 2 o: MA "i cc aUJ IT o o 25 30 TIME (MIN) I8B 180 I 3 5 6 6A 9 10 II 12 16 I8A ISC 19 20 21 2222A x2 x8 x64 xl6 x2 12x8x2 x4 x2 x2 25 30 TIME (MIN) Figure 9. Analysis of the headspace volatiles present in an aqueous extract of individually quick frozen (IQF) strawberries using a Carbowax column; top, underripe (UR) fruit; middle, normal good quality (NGQ) fruit, and bottom, overripe (OR) fruit. 66 Table 10. Normalized peak areas of TRIS volatiles and comparison of different maturities. 2 Mean normalized peak area (cm ) Comparison of means Peak Maturity level of number UR NGQ OR NGQ-UR OR-NGQ significance 0 4.80 5.16 4.91 0.36 - 0.25 0A 2.98 3.42 2.89 0.44 - 0.53 OB 3.91 4.76 4.22 0.85 - 0.54 1 10.12 10.60 11.81 0.48 1.21 2 48.23 92.16 86.53 43. 93* - 5.63 * 3 0.52 0.45 1.69 - 0.07 1.24* * 4 1.08 1.50 2.10 0.42* 0.60* * 5 0.68 0.72 0.74 0.04 0.02 6 28.40 129.60 450.21 101.20** 320. 61** ** 6A 0.87 0.92 0.91 0.05 - 0.01 7 1.15 0.80 1.81 - 0.35 1.01* * 7A 0.86 2.34 2.27 1.48* - 0.07 * 8 250. 63 284.48 340.78 33.85 56.00* * 8A 1.67 1.98 1.71 0.31 - 0.27 8B 1.98 1.95 606.10 - 0.03 604.15** ** 9 3450.62 3502.08 4020.18 51.46 518.10* * 10 286. 53 459.96 470.83 173. 43** 10.87 ** 10A 18.50 23.63 2.58 5.13 - 21.05** ** 11 32.69 38.48 78.08 5.79 39,60* * 12 18. 93a 21.06a 23.lla 2.13 2.05 12A 10.02 11.88 12. 45 1.86 0.57 13 23.08 28.56 29.51 5:48 0.95 14 18.62 16.40 17.06 - 2.22 0.66 15 102. 48 117.60 141. 18 15.12 23.58 ISA 9.40 44.16 20.83 34. 76** - 23.33* ** 16 36.72 30.90 29.81 - 5.82 - 1.09 17 16.92 18.43 18.02 1.51 - 0.41 17A 17.01 14.25 13.80 - 2.76 - 0.45 18 12.89 11.02 10.51 - 1.87 - 0.51 ISA 12.64 11.25 10.02 - 1.39 - 1.23 19 28.93 29.40 27.37 0.47 - 2.03 a. Peak area represents 0. 5 ppm of internal standard (2-butanone). *P<0.05 **P <0.01 67 Table 11. Normalized peak areas of Carbowax volatiles and comparison of different maturities. 2 Mean normalized peak area (cm ) Comparison of means Peak Maturity level of number UR NGQ OR NGQ-UR OR-NGQ significance 6 842.18 860.16 1001.24 17.98 141.08* * 6A 171.28 132.16 402.17 - 39.12 270.01** ** 9 8.70 48.96 42.29 40.26** - 6.67 ** 10 57.28 221. 76 178. 73 : 164. 48** - 43. 03 ** 11 64.21 217.60 185.22 153. 39** - 32.38 ** 12 7.98 9.36 11.08 .1.38 1.72 14 15.81 6.95 7.81 - 8.86* 0.86 * 16 38.78 41.28 46.21 2,50 4.93 18 3.71 4.85 4.90 1.14 0.05 ISA 11.71 12.90 13.47 1.19 0.57 18B 20.09 22.80 26.71 2.71 3.91 18C 7.84 6.45 13.58 - 1.39 7.13* * 18D 34.20 37.24 33.71 3.04 - 3.53 19 21.08 19.60 20.74 - 1.48 1.14 20 23. 98a 28.60a 32.61a 4.62 4.01 21 1.28 5.38 10.08 4.10* 4.70* * 22 11.88 39.36 32.71 27. 48* - 6.65 * 22A 2.81 7.68 11.42 4.87* 3.72* * 23 4.71 10.32 18.22 5.61* 7.90* * 25 10.42 13.80 9.85 3.38 - 3.95 26 8.40 10.45 9.80 2.05 - 0.65 26A 57.23 8.72 40.81 - 48.51** - 32.09** ** 27 13.71 10.62 15.82 - 3.09 5.20 28 3.50 5.45 4.17 1.95 - 1.28 29A 1.97 2.38 2.42 0.41 0.04 30 1.85 2.17 3.39 0.32 1.22 32 8.81 10.60 11.08 1.79 0.48 34 20.02 21.70 25.71 1.68 4.01 35 4.79 5.40 4.85 0.61 - 0.55 36 7.91 110.88 160. 42 102. 97** 49. 54* ** 37 6.85 8.64 9.42 1.79 0.78 39A 162. 71 146.00 170. 83 - 16.71 24.83 Peak area represents 0. 1 ppm of internal standard (n-heptanol). *P <0.05 **P <0. 01 Table 12. GLC-MS identities of TRIS peaks varying with maturity. Peak MS MS GLC Compound number identification reference confirmation 2 Propionic acid + A.S.T.M. , 1969 + 3 Isobutyric acid + ibid. + 4 n-Butyric acid + ibid. + 6 Acetaldehyde + ibid. + 7 Ethyl formate + ibid. + 7A Isobutyraldehyde + ibid. + 8 Methyl acetate + ibid. + 8B 1, 1-Dimethyoxymethane + McFadden et al. , 1964 + 9 Ethyl acetate + A.S.T.M., 1969 + 10 Ethanol + ibid. + 10A 1, 1-Diethoxymethane + McFadden et al. , 1964 + 11 Ethyl propionate + A.S.T.M., 1969 + 15A Is 1-Diethoxyethane + McFadden et al. , 1964 + oo Table 13. GLC-MS identities of Carbowax peaks varying with maturity. Peak MS MS GLC Compound number identification re ference confirmation 6 Ethyl butyrate + A. S. T.. M. , 1969 + 6A Benzaldehyde + ibid. + 9 Acetophenone + ibid. + 10 Propiophenone + ibid. 11 Propylphenylketone + ibid. 14 3- Phenylpropan-2-one + ibid. 18C Unknown 21 Ethylbenzaldehyde + ibid. 22 n-Pentyl n-hexanoate + ibid. 22A (3 - Phenylethyl acetate + ibid. + 23 cis-3-Hexen- 1-yl hexanoate + ibid. 26A 1, 1-Diethoxyoctane + McFadden et al. , 1964 36 Benzyl acetate + A. S. T. . M. , 1969 + 70 Esters The esters which changed significantly (P < 0. 05) increased with maturity. They were as follows; methyl acetate, ethyl formate, ethyl acetate, ethyl propionate, ethyl butyrate, n-pentyl -h-hexanoate, (3 -phenylethyl acetate, cis-3-hexen- 1 -yl hexanoate, and benzyl acetate. In all cases except n-pentyl n-hexanoate they increased in OR fruit as compared to NGQ fruit. The last four were present in significantly smaller amounts in UR fruit than in NGQ fruit. Increase in the level of esters on ripening has been reported previously in several products; pears (Heinz et ah , 1965; Jennings, 1967; Ton, 1968); peaches (Do et al. , 1969); bananas (Quast and Wick, 1970); and tomatoes (Dalai et al. , 1968). Increase in methyl acetate and ethyl acetate with ripening has been reported for pears (Romani and Ku, 1968) and bananas (Hultin and Proctor, 1961). Brown et al. (1966) found that ethyl propionate and ethyl butyrate in apples increased with maturity. The esters of fruits would be expected to be formed by the interaction of the corresponding alchols and acids. Such was suggested by Creveling and Jennings (1970) as a result of their work with Bartlett pear volatiles. Nursten (1970) proposed that the straight- forward reaction between an alcohol and an acid was significant in fruits only under unusual conditions such as low pH and high concentra- tions of the reactants and proposed the following key equation: 71 RCOS Co A + R'OH ^ RCOOR' + CoASH. This proposal was substantiated by the predominance in fruits of esters of even-numbered fatty acids obtained by leakage of acyl- coenzyme A intermediates from lipid synthesis and alcoholysis of these compounds (Nursten, 1970). Ester formation was proposed to be unspecific, depending only on the availability of alcohols and co- enzyme A derivatives, thus accounting for the occurrence in fruits of esters comprising all combinations of the alcohols and acids found therein (Nursten, 1970). Aldehydes The aldehydes which changed significantly (P < 0. 05) with maturity generally increased in level on ripening. Acetaldehyde, isobutyraldehyde, and ethylbenzaldehyde increased significantly from UR fruit to NGQ fruit. Acetaldehyde and benzaldehyde increased significantly in OR fruit as compared to NGQ fruit. Acetaldehyde has been reported to increase with maturity in pears (Tindale et al. , 1938; Romani and Ku, 1966), tomatoes (Dalai et al. , 1968) and carrots (Heatherbell and Wrolstad, 1971). Benzaldehyde was found to increase with maturity in peaches (Do et al. , 1969), and in tomatoes (Dalai etal. , 1968). Luh et al. (1955) found that total carbonyls increased as pears ripened. 72 Nursten (1970) suggested that at the relatively acidic pH of fruits, aldehydes could be formed to a small extent by the action of an alcohohNAD oxidoreductase enzyme system on alcohols but that largely one must look elsewhere for the origin of fruit aldehydes. Kazeniac and Hall (1970) reported that lipid-oxidizing enzymes in tomatoes acted on fatty acids to produce hydroperoxides, which decomposed to yield carbonyl compounds. Nursten (1970) stated that amino acids maybe the precursors for some fruit aldehydes, either via the a-keto acids or via a non-enzymatic Strecker reaction with suitableot-dicarbonyl compounds. Kazeniac and Hall (1970) proposed that hydroperoxides formed from fatty acids might react with amino acids to produce aldehydes. Using tracer studies on ripening tomatoes, Yu (1968) found that alanine yielded propanal and other unidentified carbonyls and that their reaction was a transamination coupled to glutamic acid; glutamine. Zenk (1966) proposed that benzaldehyde might be formed from cinnamic acid via benzoyl-S CoA by (3 -oxidation in the process of fruit ripening. Acetals The acetals which changed significantly (P < 0. 05) on ripening in this study showed rather sporadic changes with maturity. The changes were as follows: significantly lower in OR fruit than in NGQ fruit -- 1, 1-diethoxymethane, 1, 1-diethoxyethane, and 73 1, 1-diethoxyoctane; significantly higher in OR fruit than in NGQ fruit -■ 1, 1 - dimethoxymethane; significantly lower in UR fruit than in NGQ fruit -- 1, 1-diethoxyethane; significantly higher in UR fruit than in NGQ fruit -- 1, 1-diethoxyoctane. Nursten (1970) suggested that acetals in fruits could be artifacts of the treatment of the sample, particularly disruption of its cellular integrity; he proposed that acetals are formed from the aldehydes and alcohols present in fruits as their local concentrations increase. Acids The acids which changed significantly (P < 0. 05) increased with maturity. Propionic acid and n-butyric acid were present to a significantly greater extent in NGQ fruit than in UR fruit. Isobutyric acid and ri-butyric acid were present in significantly greater amounts in OR fruit than in NGQ fruit. Little work has been reported on the change in level of individual volatile acids with the ripening of fruits. Hultin and Proctor (196l) did find that acetic acid increased with ripening in bananas. Ulrich (1970) has recently reviewed the formation of acids in fruits. Aromatic Ketones The aromatic ketones that were altered significantly (P < 0. 05) in level by maturation generally increased when the fruit progressed 74 from the UR stage to the NGQ stage and then did not change significantly when fruit ripened to the OR stage. This was the case for aceto- phenone, propiophenone, and propylphenyl ketone. The exception was 3-phenylpropan-2-one, which was present to a significantly greater extent in UR fruit than in NGQ fruit. The only previous report of the alteration of such compounds in the ripening of fruit was that of Hultin and Proctor (1961). They found that two relatively high molecular weight carbonyls showed slight increases with ripening in bananas but in general stayed relatively constant. Kazeniac and Hall (1970) proposed that terpene-related ketones in tomatoes might be formed by the attack of hydroperoxides (formed by the oxidation of fatty acids by lipid-oxidizing enzymes) on terpenes such as lycopene, (3-carotene, squalene, farnesal, geranial, geraniol, and linalool. Alcohols Ethanol was the only alcohol which changed significantly (P < 0. 05) with maturity. It increased with the change from UR to NGQ fruit but was not significantly altered with the change from NGQ fruit to OR fruit. Ethanol has been found to increase on ripening in bananas (Hultin and Proctor, 1961; Murray £tal. , 1968), pears (Tindale et al. , 1938), and carrots (Heatherbell and Wrolstad, 1971). 75 However, Ton (1968) found that ethanol stayed constant throughout the ripening of pears. Nursten (1970) stated that ethanol was generally formed by reactions analogous to anaerobic fermentation; that is, via pyruvic acid and acetaldehyde. Ethanol could be formed from alanine via a transamination reaction, probably linked to glutamic acid:glutamine (ibid. ). Other Considerations One peak which was not identified (Carbowax peak 18C) was present in a significantly greater amount (P < 0.05) in OR fruit than in NGQ fruit. In all cases its mass spectrum was so weak that not even a tentative identification could be advanced. The results of this study are unique in that there were lower boilers (TRIS peaks) and higher boilers (Carbowax peaks) that increased significantly (P < 0. 05) with maturity. Mehlitz and Gierschner (1962) found that lower boilers increased and higher boilers decreased on ripening in apples. Heatherbell and Wrolstad (1971) found that lower boilers increased and higher boilers stayed relatively constant with maturation of carrots. No peak changing significantly with maturity was identified as ethylene. This would be in accordance with the work of Rhodes (1970), who classified strawberries as a non-climacteric fruit. 76 It must be emphasized that a vapor analysis technique such as the one used in this research allows one to say only that this method shov/ed or failed to show significant changes in volatiles with ripening. Myers et al. (1969) found that GLC vapor analysis showed isoamyl alcohol and isoamyl acetate to be constant during ripening of bananas but that tracer studies showed these compounds to be continually produced from labelled leucine during maturation. Such a situation could have prevailed in the case of this study, limiting* the force with which conclusions can be stated. 77 BIBLIOGRAPHY Ahmed, E. M. and L. E. Scott. 1963. Gas-liquid chromatography studies with strawberries. Proceedings of the American Society for Horticultural Science 82;299-310. Albersheim, P. 1965. Biogenesis of the cell wall. In; Plant biochemistry, ed. J. Bonner and J. E. Varner. New York, Academic Press, p. 298-321. Alexander, M. 1970. Gas chromatography for detection and charac^ terization of microorganisms. Food Technology 24:372-376. Amerine, M. A. and W. V. Cruess. i960. The technology of wine making. Westport, Connecticut, Avi Publishing Company, p. 74-75. Angelini, P. and I. J. Pflug. 1967. Volatiles in controlled- atmosphere apple storage: evaluation by gas chromatography and mass spectrometry. Food Technology 2 1:99-102. Anonymous. 1970. Molds on foods. FDA Papers 4(8):2l-24. A. S. T. M. 1969. American Society for Testing and Materials. Index of mass spectral data. Philadelphia, Pennsylvania, p. 1-624. Barash, I. , J. M. Klisiewicz and T. Kosuge. 1964. Biochemical factors affecting pathogenicity of Botrytis cinerea on safflower. Phytopathology 54:923-927. Bassette, R. , R. E. Bawdon and T. J. Claydon. 1967. Production of volatile materials in milk by some species of bacteria. Journal of Dairy Science 50:167-171. Bills, D. D. and T. W. Kennan. 1968. Dimethyl sulfide and its precursor in sweet corn. Journal of Agricultural and Food Chemistry 16:643-645. Birkinshaw, J. H. , H. Raistrick and G. Smith. 1942. Studies on the biochemistry of micro-organisms. Fumaryl-dl-alanine (fumaromono-dl-alanide), a metabolic product of Penicillium resticulosum sp. nov. The Biochemical Journal 36:829-835. Black, D. R. e* al. 1971. Strawberry volatiles from different varieties. Paper no. 8, Agricultural and Food Chemistry Division, l6lst American Chemical Society National Meeting. March 28-April 2. Los Angeles, California. 78 Brown, D. S. , J. R. Buchanan and J. R. Hicks. 1966. Volatiles from apple fruit as related'to variety, maturity, and ripeness. Proceedings of the American Society for Horticultural Science 88:98-104. Campbell, A. D. 1970. Food and Drug Administration, Falls Church, Virginia. Personal communication. Charpentie, Y. 1954. Factors in wine acidity. Annales de I'Institut National de la Recherche Agronomiqu-e 3:89- 167. (Abstracted in Chemical Abstracts 5 l:no. 17,084i. 1957) Chaudhary, S. S. , A. D. Webb and R. E. Kepner. 1968. GLC investigation of the volatile compounds in extracts from Sauvignon blanc wines from normal and Botrytised grapes. American Journal of Enology and Viticulture 19:6-12. Cochrane, V. W. 1958. Physiology of fungi. New York, Wiley. p. 132-134. Coppens, A. and L. Hoejenbos. 1939. Investigation of the volatile constituents of strawberry juice (Fragaria eliator Ehrh. ). Recueil des Travaux Chimiques des Pays-Bas 58:680-690. Corse, J. and K. P. Dimick. 1958. Volatile flavors of strawberry. In: Flavor research and food acceptance, ed. A. D. Little. New York, Reinhold. p. 302-314. Creveling, R. K. and W. G. Jennings. 1970. Volatile components of Bartlett pear. Higher boiling esters. Journal of Agricultural and Food Chemistry 18:19-24. Dalai, K. B. et al. 1968. Volatile components of developing tomato fruit grown under field and greenhouse conditions. Plant and Cell Physiology 9:389-400. Day, E. A. 1967. Cheese flavor. In: Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schultz, E. A. Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 331-361. De Jong, D. W. , A. D. King, Jr. and F. P. Boyle. 1968. Modifica- tion of white table wines with enzymes from Botrytis cinerea Pers. American Journal of Enology and Viticulture 19:228-237. 79 Difco Laboratories. 1953. Difco manual of dehydrated culture media and reagents for microbiological and clinical laboratory procedures. Ninth ed. Detroit, Michigan, Difco Laboratories, Inc. p. 239. Dimick, K. P. and B. B. Makower. 1956. Volatile flavor of straw- berry essence. I. Identification of the carbonyls and certain low-boiling substances. Food Technology 10;73-75. Dimick, K. P. and J. Corse. 1958. The volatile flavors of straw- berry. American Perfumer and Aromatics 71:45, 48, 53. Dittrick, H. 1964. Glycerin formation by B_. cinerea in grapes and musts as well as the glycerin content of berry-selected and dry-berry-selected wines. Wein-Wissenschaft 19:12-20. (Abstracted in American Journal of Enology and Viticulture 15:222. 1964) Do, J. Y. , D. K. Salunkhe and L. E. Olson. 1969. Isolation, identification and comparison of the volatiles of peach fruit as related to harvest maturity and artificial ripening. Journal of Food Science 34:618-621. Dougherty, J. D. , R. D. Campbell and R. L. Morris. 1966. Actinomycete: isolation and identification of agent responsible for musty odors. Science 152:1372-1373. Dravnieks, A. and A. O'Donnell. 1970. Principles and some techniques of high-resolution headspace analysis. Paper no. 2, Agricultural and Food Chemistry Division, l60th American Chemical Society National Meeting, September 14-18. Chicago, Illinois. Eisenberg, W. V. 1969. Report on analytical mycology of food and drugs. Journal of the Association of Official Analytical Chemists 52:279-280. Fields, M. L. and B. Richmond. 1967. Effect of pH on fungal metabolites used as indicators of quality of foods. Journal of Food Science 32:235-237. Fields, M. L. , B. S. Richmond and R. E. Baldwin. 1968. Food quality as determined by metabolic by-products of microorgan- isms. In: Advances in Food Research, ed. C. O. Chichester, E. M. Mrak and G. F. Stewart. New York, Academic Press, p. 161-229. 80 Filipic, V. J. , J. C. Underwood and C. J. Dooley. 1969. Trace components of the flavor fraction of maple syrup. Journal of Food Science 34; 105- 110. Franke, W. , G. Jilge and G. Eichhorn. 1961. Enzymic transforma- tion of C2-acids by microorganisms. I. Glyoxylate trans- aminases from fungi, especially Aspergillus niger. Archiv fur Mikrobiologie 39:58-87. Furia, T. E. (ed. ). 1968. Handbook of food additives. Cleveland, Ohio, Chemical Rubber Company, p. 169-172, 248, 252, 484, 567, 592, 661. Gerber, N. N. and H. A. Lechevalier. 1965. Geosmin, an earthy- smelling substance isolated from actinomycetes. Applied Microbiology 13:935-938. Giannone, L. and M. Tommasicchio. 1969. A search for a chemical method in place of the Howard method for tomato products. Industria Conserve 44:21-24. (Abstracted in Food Science and Technology Abstracts 2:no. 75-699. 1970) Gilman, J. C. 1957. A manual of soil fungi. Ames. Iowa State University Press, p. 1-11, 299, 300. Gray, William D. 1959. The relation of fungi to human affairs. New York, Holt. p. 294-296. Gray, W. D. 1970. The use of fungi as food and in food processing. Cleveland, Ohio, Chemical Rubber Company, p. 90-100. Gross, G. G. , K. H. Bolleart and M. H. Zenk. 1968. Reduction of cinnamic acid to cinnamaldehyde and alcohol. Biochemical and Biophysical Research Communications 322 173-178. Guadagni, D. G. , J. C. Miers and D. Venstrom. 1968. Methyl sulfide concentration, odor intensity, and aroma quality in canned tomato juice. Food Technology 22s 1003- 1006. Guarino, P. A. and A. Kramer. 1969. Gas chromatographic analysis of head-space vapors to identify micro-organisms in foods. Journal of Food Science 34:31-37. Gumbmann, M. R. and H. K. Burr. 1964. Volatile sulphur compounds in potatoes. Journal of Agricultural and Food Chemistry 12:404-408. Hancock, J. G. , R. L. Millar and J. W. Lorbeer. 1964. Pectolytic and cellulolytic enzymes produced by Botrytis allii, B. cinerea, and B_. squamosa in vitro and in vivo. Phytopathology 54:928-931. 81 Hansen, A. P. and P. G. Keeney. 1970. Comparison of carbonyl compounds in moldy and non-moldy cocoa beans. Journal of Food Science 35;37-40. Harvey, J. M. and W. T. Pentzer. 1966. Market diseases of grapes and other small fruits. Agriculture Handbook no. 189, Market Quality Research Division, Agricultural Research Service, United States Department of Agriculture, Washington, D. C. Heatherbell, D. A., R. E. Wrolstad and L. M. Libbey. 1970. Isolation, concentration and analysis of carrot volatiles using on- column trapping and GLC-MS. Paper no. 21, Agricultural and Food Chemistry Division, l60th American Chemical Society National Meeting, September 14-18. Chicago, Illinois. 1971. Carrot volatiles. 1. Characterization and effects of canning and freeze drying. Journal of Food Science 36:219-224. Heatherbell, D. A. and R. E. Wrolstad. 1971. Carrot volatiles. 2. Influence of variety, maturity, and storage. Journal of Food Science 36;225-227. Heinz, D. E. , R. K. Creveling and W. G. Jennings. 1965. Direct determination of aroma compounds as an index of pear maturity. Journal of Food Science 30s641-643. Henderson, M. E. K. and V. C. Farmer. 1955. Utilization by soil fungi of p-hydroxybenzaldehyde, ferulic acid, syringaldehyde, and vanillin. Journal of General Microbiology l2;37-46. Henis, W. , J. R. Gould and M. Alexander. 1966. Detection and identification of bacteria by gas chromatography. Applied Microbiology 14;513-524. Hodge, J. E. 1967. Nonenzymatic browning reactions. In; Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schuitz, E. A. Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 465-491. Horwitz, W. (ed. ). 1965. Official methods of analysis of Associa- tion of Official Agricultural Chemists. 10th ed. Washington, D. C. , Association of Official Agricultural Chemists, p. 722- 724. 82 Hultin, H. O. and B. E. Proctor. 1961. Changes in some volatile constituents of the banana during ripening, storage, and pro- cessing. Food Technology 15S440-444. Jarvis, W. R. 1964. Effect of some climatic factors on the incidence of grey mould of strawberry and raspberry fruit. Horticultural Research 3:65-71. Jennings, W. G. 1967. Peaches and pears. Ins Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schultz, E. A. Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 419-430. 1970. A comparison of sampling techniques for GLC. Paper no. 99, Agricultural and Food Chemistry Division, 160th American Chemical Society National Meeting, September 14-18. Chicago, Illinois. Jennings, W. G. and M. R. Sevenants. 1964. Volatile components of peach. Journal of Food Science 29:796-801. Kaminski, E. , S. Stawicki and A. Niewiarowicz. 1969. The origin and characterization of mustiness in grain. Technical Report 3, Division of Food Biochemistry and Microbiology, Department of Agricultural and Food Technoldgy, Agricultural University of Poznan, Poland. Katayama, O. , Y. Watanabe and K. Tsubata. 1968. Aroma compo- nents in fruits and vegetables. IV. Strawberry volatiles and their changes in jam processing. Nippon Shokuhin Kogyo Gakkaishi 15:8- 17. (Abstracted in Chemical Abstracts 69: no. 75, 726z. 1968) Katayama, O. , K. Tsubata and I. Yamato. 1967. Aroma components in fruits and vegetables. II. Volatile components of tomato. Nippon Shokuhin Kogyo Gakkaishi 14:444-449. (Abstracted in Chemical Abstracts 68:no. 103, 982e. 1968) Kazeniac, S. J. and R. M. Hall. 1970. Flavor chemistry of tomato volatiles. Journal of Food Science 35:519-530. Keenan, T. W. and D. D. Bills. 1968. Metabolism of volatile compounds by lactic starter culture microorganisms. A review. Journal of Dairy Science 51:1561-1567. 83 Keenan, T, W. and R, C, Lindsay. 1968. Evidence for a dimethyl sulfide precursor in milk. Journal of Dairy Science 51:112- 114. King, A. D. , Jr., W. M. Camirand and K. L, Mihara. 1969. Submerged-culture production of Botrytis cinerea mycelium. American Journal of Enology and Viticulture 20s 146-151. Kingston, B. H. 1964. The interpretation and value of gas liquid chromatography (G. L. C. ) data. American Perfumer and Cosmetics 79(11 ):50-52. Kretovich, V. L, and A. P. Prokhovora. I960. Biochemical characteristics of grains possessing different odors. Izvestiza Akademic Nauk Azerbaidzhanskoi Soyuz Sovetskikh Sotsialistic- heskikh Respublik, Seriya Biologicheskikh. p. 446-450. (Abstracted in Chemical Abstracts 67;no. 63, 084j. 1967) Libbey, L. M. 1971. Department of Food Science and Technology. Corvallis, Oregon State University. Personal communication. Liener, I. E. 1966. Cyanogenetic glycosides. In: Toxicants occurring naturally in foods. Food Protection Committee, Food and Nutrition Board, National Academy of Sciences, National Research Council. Washington, D. C. , National Academy of Sciences, Printing and Publishing Office, p. 58-61, Lindsay, R. C. 1967. Cultured dairy products. In: Symposium on Foods: The chemistry and physiology of flavors, ed. H, W. Schultz, E, A. Day and L. M, Libbey, Westport, Connecticut, Avi Publishing Company, p. 315-330. Luh, B. S. et al. 1955. Volatile reducing substances in canned Bartlett pears. Food Technology 9:639-642. MacLoed, A. J. and G. MacLoed, 1970. Flavor volatiles of some cooked vegetables. Journal of Food Science 35:734-738. Marion, J. P. et al. 1967. Sur la composition de I'ar'Sme de cacao. Helvetica Chimica Acta 50:1509-1516, Matthews, R. F. i960. Gas and paper chromatography of volatile flavor constituents of several vegetables. Ph. D. thesis. Ithaca, New York, Cornell University. 84 McFadden, W. H, et al. 1964. Correlations and anomalies in mass spectra: acetals. Analytical Chemistry 36:1031-1037. McFadden, W. H. et al. 1965. Volatiles from strawberries. II. Combined mass spectrometry and gas chromatography on complex mixtures. Journal of Chromatography 18:10-19. Mehlitz, A. and K. Gierschner. 1962. Further investigations on aroma concentrates from fruit juices. International Fruchtsaft Union Bericht Wissenschaft Technikisch Kommunikation no. 4:301-318. (Abstracted in Chemical Abstracts 58:no. 9562f. 1963) Minarik, E. 1962. Utilization of molds and enzyme preparations prepared from them for the production of grape wines. Vinarstvi 54:167-169. (Abstracted in Chemical Abstracts 56: no. 7799b. 1962) Mishra, J. N. 1953. A study of the nutritional requirement of Botrytis cinerea Pers. Bihar Academy of Agricultural Science Proceedings 2:81 - 90. Mitruka, B. M. and M. Alexander. 1968. Gas chromatography for detection of viral infections. Science 160:309-311. Mitruka, B. M. , N. L. Norcross and M. Alexander. 1968. Gas chromatographic detection of in vivo activity of equine infec- tious anaemia virus. Applied Microbiology 16:1093-1094. Moorhouse, B. R. and H. Salwin. 1969. Investigation of volatile reducing substances as an indicator of decomposition in raw and processed foods. Journal of the Association of Official Analytical Chemists 52:1135-1141. Morgan, M. E. and E. A. Day. 1965. Simple on-column trapping procedure for gas chromatographic analysis of flavor volatiles. Journal of Dairy Science 48:1382-1384. Morgan, M. E. , S. G. Kayser and R. A. Scanlan. 1971. Department of Food Science and Technology. Corvallis, Oregon State University. Personal communication. Motoc, D. and S. C. Brad. 1963. The enzymic complex in molds developing on grapes. Industria Alimentara, Produse Vegetable 14:385-389. (Abstracted in Chemical Abstracts 60:no. 11,076c. 1964) 85 Murray, K. E. et al. 1968. The volatile alcohols of ripe bananas. Journal of Food Science 33:632-634. Myers, M. J. , P. Issenberg and E. L. Wick. 1969. Vapor analysis of the production by banana fruit of certain volatile constituents. Journal of Food Science 34:504-509. Nambudiri, A. M. D. , P. V. Subba Rao and J. V. Bhat. 1970. Metabolism of aromatic compounds by an Alternaria species. Phytochemistry 9:687-693. Nelson, K. E. and M. A. Amerine. 1956. Use of Botrytis cinerea for the production of sweet table wines. American Journal of Enology 7:131-136. Nelson, P. E. and J. E. Hoff. 1969. Tomato volatiles: effect of variety, processing, and storage time. Journal of Food Science 34:53-57. Niewiadomski, H. and J. Salmonowicz. 1969. Metabolism of fatty acids by molds and basic studies of biological production of essential fatty acids. Vitalstoffe-Zivilisationskrankheiten 14: 146-148. (Abstracted in Food Science and Technology Abstracts 2:no. 2N-32. 1970) Novak, E. K. and G. Voros-Felaki. 1958. Carbon metabolism in Botryotinia fuckeliana (Botrytis cinerea) and its bearing on sweet rot in grapes. I. Organic acids, the only carbon sources of the mold. Acta Microgiologica Academiae Scientiarum Hungaricae 5:217-221. (Abstracted in Chemical Abstracts 54:no. 14, 359b. 1960) Nursten, H. E. 1970. Volatile compounds: the aroma of fruits. In: The biochemistry of fruits and their products, ed. A. C. Hulme. New York, Academic Press, p. 239-268. Nursten, H. E. and A. A. Williams. 1967. Fruit aromas: a survey of components identified. Chemistry and Industry, p. 486-497. Ovcharov, K. 1937. Production of thiourea by fungi. Comptes Rendus de rAcademie des Sciences de 1'U. R.S. S. 16:461-464. (Abstracted in Chemical Abstracts 32:no. 1745(7). 1938) Parks, O. W. 1967. Milk flavor. In: Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schultz, E. A. 86 Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 296-314. Parks, O. W. and S. Patton. 196l. Volatile carbonyl compounds in stored dry whole milk. Journal of Dairy Science 44; 1-9. Patterson, W. I. 1945. Report on decomposition in foods. Journal of the Association of Official Agricultural Chemists 28:233-239. Pelczar, M. J. , Jr. 1957. Manual of microbiological methods. New York, McGraw-Hill. p. 118. Pitt, D. 1968. Histochemical demonstration of certain hydrolytic enzymes within cytoplasmic particles of Botrytis cinerea. Journal of General Microbiology 52.»67-75. Ponting, J. D. , B. Feinberg and F. P. Boyle. 1968. Fruits: characteristics and the stability of the frozen product. In: The freezing preservation of foods. Vol. 2. Factors affecting quality in frozen foods, ed. D. K. Tressler, W. B. Van Arsdel and M. J. Copley. "Westport, Connecticut, Avi Publishing Company, p. 107-133. Popova, E. M. and M. G. Puchkova. 195 3. Treatment of champagne stock with enzyme preparation in the vat method. Biokhimica Vinodeliya Sbornik 4:153-160. (Abstracted in Chemical Abstracts 49:no. 56ld. 1955) Powelson, R. L. I960. Initiation of strawberry fruit rot caused by Botrytis cinerea. Phytopathology 50:491-494. Quast, S. V. and E. L. Wick. 1970. Volatile constituents of ripen- ing banana fruit. Paper no. 74, Agricultural and Food Chem- istry Division, l60th American Chemical Society National Meeting, September 14-18. Chicago, Illinois. Reddy, M. C. et al. 1968. Ester production by Pseudomonas fragi. I. Identification and quantification of some esters produced in milk cultures. Journal of Dairy Science 51:656-659. Reiner, E. 1965. Identification of bacterial strains by pyrolysis- gas-liquid chromatography. Nature 206:1272-1274. Rhodes, M. J. C. 1970. The climacteric and ripening of fruits. In: The biochemistry of fruits and their products, ed. A. C. Hulme. New York, Academic Press, p. 521-533. Rohan, T. A. and T. Stewart. 1967. The precursors of chocolate 87 aroma: production of free amino acids during fermentation of cocoa beans. Journal of Food Science 32:395-398. Romani, R. J. and L. L. Ku. 1966. Direct gas chromatographic analysis of volatiles produced by ripening pears. Journal of Food Science 31:558-560. Romano, A. H. and R. S. Safferman. 1963. Studies on actinomycetes and their odors. American Water Works Association Journal 55:169-176. Rosen, A. A. et_ al. 1968. Identity of odorous substance produced by Streptomyces griseoluteus. Applied Microbiology 16:178-179. Rubin, B. A. and T, M. Ivanova. 1958. Oxidative transformation of amino acids upon interaction of cabbage tissue with the fungus Botrytis cinerea. Biokhimiya 23:540-546. Salkova, E. G. , R. I. Bekbutatova and K. H. Ngo. 1966. Role of parasite pectolytic enzymes and host plant phenols in the development of infection. Trudy Vserossiiskogo Nauchno- Isseldovatel'sleogo Khimicheskogo Instituta Promyshlennosti Mestnogo Podchineniya 26:277-231. (Abstracted in Chemical Abstracts 69:no. 65,207s. 1968) Salwin, H. 1969. Report on decomposition and filth in foods (chemical indexes). Journal of the Association of Official Analytical Chemists 52:243-244. 1970. Report on decomposition and filth in foods (chemical indexes). Journal of the Association of Official Analytical Chemists 53:323-325. Sandoval, A. and H. Salwin. 1968. Effects of decomposition and freeze-drying on gas chromatographic analysis of headspace vapors of beef and strawberries. Journal of the Association of Official Analytical Chemists 51:1210-1212. Saravacos, G. D. and J. C. Moyer. 1968. Volatility of some flavor compounds during freeze-drying of foods. Chemical Engineer- ing Progress, Symposium Series 64(86):37-42. Scanlan, R. A. et al. 1968. Heat-induced volatile compounds in milk. Journal of Dairy Science 51:1001-1007. Seidel, C. F. , H. Schinz and M. Stoll. 1958. Compounds with aroma. II. Strawberry aroma. Helveta Chimica Acta 41:372-377. 88 Self, R. 1967. Potato flavor. In; Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schultz, E. A. Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 362-389. Self, R. , J. C. Casey and T. Swain. 1963. The low-boiling volatiles of cooked foods. Chemistry and Industry, p. 863, 864. Self, R. , H. L. J. Rolley and A. E. Joyce. 1963. Volatile compounds from cooked potatoes. Journal of the Science of Food and Agriculture 1428-14. Sloan, J. L. , D. D. Bills and L.. M. Libbey. 1969. Heat-induced compounds in strawberries. Journal of Agricultural and Food Chemistry 17:1370-1372. Sprenger Institute. 1969. Annual report of research on storage and processing of horticultural produce. Wageningen, The Nether- lands, p. 134, 135. Steel, R. G. D. and J. H. Torrie. I960. Principles and procedures of statistics. New York, McGraw-Hill. p. 436-441, 453. Stevens, K. L. et al. 1969. Volatiles from grapes. Comparison of Grenache juice and Grenache rose wine. Journal of Agricultural and Food Chemistry 17:1102-1106. Stevens, M. A. 1971. Department of Vegetable Crops. Davis, University of California. Personal communication. Stevens, M. A. _et al. 1967. Volatile components of canned snap beans (Phaseolus vulgaris L. ). Proceedings of the American Society for Horticultural Science 91:833-845. Strackenbrock, K. H. 1962. Gas chromatographic measurement of aroma quality of fruit juices as influenced by various technological methods. International Fruchtsaft Union Bericht Technikisch Kommunikation No. 4:287-299. (Abstracted in Chemical Abstracts 58:no. 9562e. 1963) Sunkist Growers. 1964. Preservers handbook. Ontario, California, Sunkist Growers Inc. p. 31-35. 89 Tagawa, K. and A. Kaji. 1967. Pectic enzymes. XXIV. Production of pectic enzymes by Botryti's cinerea and observations on plant tissue maceration. Hakko Kogaku Zasshi 45:347-352. (Abstracted in Chemical Abstracts 69:no. 42, 771g. 1968) Tanaka, A., S. Shimizu and S. Fukuh 1968. Utilization of hydro- carbons by microorganisms. IV. Growth of molds on hydro- carbons. Hakko Kogaku Zasshi 46:461-467. (Abstracted in Chemical Abstracts 69:no. 65,3l6b. 1968) Tanner, F. W. 1944. The microbiology of foods. Champaign, Illinois, Garrard Press, p. 117. Teranishi, R. et ah 1963. Volatiles from strawberries. I. Mass spectral identification of the more volatile components. Journal of Food Science 28:478-483. Tindale, G. B. , S. A. Trout and F. E. Heulin. 1938. Investigations on the storage, ripening, and respiration of pears. Journal of the Department of Agriculture of Victoria 36:34-52, 90. (Cited in: Jennings, W. G. 1967. Peaches and pears. In: Symposium on Foods: The chemistry and physiology of flavors, ed. H. W. Schultz, E. A. Day and L. M. Libbey. Westport, Connecticut, Avi Publishing Company, p. 419-430.) Ton, P. C. 1968. Composition of aroma of pears and its variation during maturation. Fruits 23:423-430. (Abstracted in Chemical Abstracts 70:no. 10,495. 1969) Tressl, R. , F. Drawert and W. Heimann. 1969. Identification of the constituents of strawberry aroma by gas chromatography and mass spectrometry. Zeitschrift ftir Naturforschung 24B:1201- 1202. U.S. Department of Agriculture, 1968. Progress report of the Market Quality Research Division, Agricultural Research Service. Hyattsville, Maryland, p. 22, 1969. Inspector's instructions for mold counting. Technical Inspection Procedures no. 135-A8. Washington, D. C. p. 7. Ulrich, R. 1970. Organic acids. In: The biochemistry of fruits and their products, ed. A. C. Hulme. New York, Academic Press, p. 239-268. 90 Usseglio-Tomasset, L. 1958. Some aspects of the metabolism of Botrytis cinerea and of Penicillium expansum growing on grape must. Annali della Sperimentazione Agraria 12:1689-1705. (Abstracted in Chemical Abstracts 53:no. 22, 230i. 1959) Varseveld, G. W. and D. V. Beavers. 1966. Preserving character- istics of Hood, Marshall, and Northwest strawberry varieties. Special report 213. CoryalliS, Oregon State University Agricultural Experiment Station. Van Den Berg, L. and S. M. Tang. 1969. Effect of relative humidity on production of extracellular pectolytic enzymes by Botrytis cinerea and Sclerotinia sclerotiorum. Canadian Journal of Botany 47:1007-1010. Walter, R. H. and I. S. Fagerson. 1968. Volatile compounds from heated glucose. Journal of Food Science 33:294-297. Watson, J. T. 1969. Gas chromatography and mass spectroscopy. In: Ancillary techniques of gas chromatography, ed. L. S. Ettre and W. H. McFadden. New York, Wiley Interscience. p. 174-178. Whiting, G. C. 1970. Sugars. In; The biochemistry of fruits and their products, ed. A. C. Hulme. New York, Academic Press, p. 1-31. Widdowson, E. M. and R. A. McCance. 1935. The available carbohydrate of fruits. Determination of glucose, fructose, sucrose, and starch. Biochemical Journal 29:151-156. Willhalm, B. , E. Palluy and M. Winter. 1966. Flavors. XII. The flavor of fresh strawberries. Identification of the volatile acids and other compounds. Helveta Chimica Acta 49:65-67. Williams, K. T. , E. F. Potter and A. Bevenue. 1952. A study, by paper chromatography, of the occurrence of nonfermentable sugars in plant materials. Journal of the Association of Official Agricultural Chemists 35:483-486. Wilson, B. J. 1966. Fungal toxins. In: Toxicants occurring naturally in foods. Washington, D. C. , National Academy of Sciences Printing and Publishing Office, p. 126-146. 91 Winstead, N. N. and J. C. Walker. 1954. Production of vascular browning by metabolites from several pathogens. Phytopathol- ogy 44;153-158. Winter, M. 1958. Volatile strawberry fruit flavor by paper and column chromatography. In: Flavor research and food acceptance, ed. A. D. Little. New York, Reinhold. p. 290-301. Winter, M. ^t al. 1958. Chromatography on columns and paper in the study of the volatile fruit flavor of strawberries. Perfumery and Essential Oil Record 49;250-255. Winter, M. and E. Sundt. 1962. Flavors. IX. Flavor of fresh strawberries. Analysis of volatile alcohols. Helveta Chimica Acta 45:2212-2221. Winter, M. and B. Willhalm. 1964. Flavors. X. Flavor of fresh strawberries. Analysis of the carbonyl compounds, esters and volatile alcohols. Helveta Chimica Acta 47:1215-1227. Wong, F. F. and J. F. Carson. 1966. Isolation of S-methyl methionine sulfonium salt from fresh tomatoes. Journal of Agricultural and Food Chemistry 14:247-249. Yu, M. H. 1968. Amino acids as precursors of volatile components in tomato fruit. Dissertation Abstracts, Section B29:1262. Zenk, M. H. 1966. Aldehydes. In: Biosynthesis of aromatic compounds, ed. G. Billek. Oxford, Pergamon Press, p. 45-60. APPENDICES 92 APPENDIX I Fortran IV Computer Program to Calculate Normalized Peak Areas for GLC Data This appendix contains a computer listing of the program, of sample preliminary information cards, and of sample cards pertain- ing to a GLC run. It also contains information on the format of all of the cards needed to use the program as well as a description of the output of the program. Computer Listing A listing of the program, a listing of sample preliminary information cards (for internal standard retention times, for internal standard peak areas, for factors for amount of filtrate used in the headspace vial, for factors for filtrate yield, and for reference peak areas for internal standards), and a listing of sample cards for GLC peaks are given. One sample card is included for each type of preliminary information. To illustrate cards pertaining to an individual GLC run the following are included; the card used at the beginning of a GLC run to cause printing of column headings and explanatory phrases in the output; a card used for the information for a GLC peak (peak OA in a TRIS run, in this case); and the card which causes total peak area to be summed and total normalized peak area to be summed at the end of the group of peak cards for a GLC run. 93 •JOB.708065.AMOS.SAVE FOR FRANK AMOS 'TIME=1000 •FORT-AN.L.R PROGRAM AMOSGC1 C THIS PROGRAM CALCULATES RELATIVE RETENTION TIME. PEAK AREA, AND C NORMALIZED PEAK AREA FOR GLC PEAKSt DIMENSION HINSTDTR(500), STDAREA(500) ♦ FAMTFILT(500!. FILYIELD<500 9). SAREAFAC(100), AREANFAC(500) . TNORMFAC(500)» TOTAREA(500). TOTN 9AREA(50C) DO 5001=1,500 5 00 HINSTDTRlI)=STDAREA(I)=FAMTFILT(I)=FILYI ELD(I)^AREANFAC(I)=TNORMFA 9C(I)=TOTAREA(I)=TOTNAREA(I 1=0 DO 5051=1.100 505 SAREAFACf I)=0 IARVARKE=0 ISPPRINT=0 ISAFCKEY=0 N = 0 EXPTID=0 COLUMNID=0 PEAKID=0 PEAKTR=0 PKHEIGHT=0 HAFWIDTH=0 ATTMULT=0 PEKRELTR=0 PEAKAREA=0 TNRMAREA=0 NTOAREAK=0 100 FORMAT(I1,1X,I1.1X,I1,5X,I3,2X.R4.1X,R5.1X.RA.1X,F4.1»1X,2F5.1»1X, 9F4.0.1X, II) 171 FORMAT(7(16F5.1/) , (8F5.1) ) 172 FORMATf 10X.7F10.1/ .14(8F10. 1 / ) , 1F10.1 ) 173 FORMAT(5X,15F5.3/»6(16F5.3/).9F5.3) 174 FORMAT(5X , 15F5.4/,6(16F5.4/ ) ,9F5.4) 175 FORMAT(4F10.1 ) 110 FORMAT!1H0.R4»4X,R5»4X,R4»3X,F7.1.3X.F7.4»3X.F8.2.2X,F14.2»2X,F10. 92,F26.2,F28.7) 200 F0RMAT<1H1,15HEXPERIMENT ID =,32X,R4) 210 FORMAT(1HO,9HCOLUMN IS,36X,R5) 220 FORMAT(1HO,34HINTERNAL STANDARD RETENTION TIME =,6X.F20.1) 230 FORMAT!1H0.3HN =,44X,I3) 240 FORMAT!1H0.25HINTERNAL STD. PEAK AREA =,1X,F24.2) 243 FORMAT(1HO,37HSTANDARD AREA FACTOR COMPARISON KEY =,1X,I1) 244 FORMAT!1H0,22HSTANDARD AREA FACTOR =.F18.1) 250 FORMAT(1HO,25HAPEA NORMALIZING FACTOR =,15X,F30.5) 260 FORMAT!1H0,38HNORMALIZING FACTOR FOR FILTRATE AMT. =.2X.F25.5) 270 FORMAT(1H0.39HNORMALIZING FACTOR FOR FILTRATE YIELD =»1X.F2!.5) 297 FORMAT!1HO»28HOVERALL NORMALIZING FACTOR =.12X,F25.5) 300 FORMAT!1H1,134HEXPT. COLUMN PEAK PEAK TR REL. TR PEAK HT 9. PEAK 1/2 WIDTH ATT. MULT. PEAK ArtEA NOR 9MALIZED PEAK AREA) 7010 FORMAT(1HO,17HTOTAL PEAK AREA =.F40.5) 7C20 FORMAT(lH0,28HTOTAL NORMALIZED PEAK AREA =,F40.5> 5 READ(60,100) IARVARKE.ISPPRI NT,ISAFCKEY,N,EXPTID.COLUMNID.PEAK ID,PE 9AKTR,PKHEIGHT,HAFWIDTH,ATTMULT,NTOAREAK GO TO(999,6)EOFCKF(60) 6 IF! IARVARKE.LE.5)70.11 70 GO TO!71,72,73,74,75)IARVARKE 71 READ!60.171)(HINSTDTR(K),K=1,120) GO TO 5 7 2 READ(60,172) (STDAREA!K),K=1,120) GO TO 5 94 73 READ<60.1 73) (FAMTFILT(K) .IC = 1,120) GO TO 5 74 READ(60.1 74)(FILYIELD(K).K=1.120) GO TO 5 75 READ(60.1 75 I (SAREAFAC(K).K = lt4) GO TO 5 11 AREANFACI N)=STDAREA(N)/SAREAFAC(ISAFCKEY) TNORMFACI N)=AREANFAC(N)*FAMTFILT(N)*FILYIELD(N) IF(ISPPRI NT.£0.9)255,89 255 WRITE(61 ,200)EXPTID WRITE(61 »210)COLUMNID WR1TE(61 .220)HINSTDTR(N) WRITE(61 ,230)N WRITE(61 .240)STDAREA(N) WRITE(61 .243 ) ISAFCKEY WRITE!61 .244)SAREAFAC(ISAFCKEY) WRITE(61 .250)AREANFAC(N) WRITE(61 .260)FAMTFILT(N) WRITE(61 ,270)FILYIELD(N) WRITE(61 ,297)TNORMFAC(N) WRITE(61 .300) GO TO 5 89 IFINTOARE AK.EQo9)G0 TO 701 PEKRELTR= PEAKTR/HINSTDTRIN) PEAKAREA= PKHEIGHT*HAFWIDTH»ATTMULT TOTAREAtN )=TOTAREA(N)+PEAKAREA TNRMAREA= PEAKAREA*TNORMFAC(N) TOTNAREAt N)=TOTNAREA(N)+TNRMAREA WRITE(61. IIOIEXPTID.COLUMNID.PEAKID.PEAKTR.PEKREUTR.PKHEIGHT. 9HAFWIDTH, ATTMULT.PEAKAREA.TNRMAREA GO TO 5 701 WRITE(61 i 200)EXPTI0 WRITE(61. 210)COLUMNID WRITE(61 . 230IN WRITE(61» 7010)TOTAREA(N) WRITE(61 , 7020)TOTNAREA(N) GO TO 5 999 STOP END t i 1 1405 1340 1405 1060 1385 1350 1390 1365 1385 1430 1395 1395 1415 1390 1395 1360 3640 3640 49500 113280 136496 145360 127600 166400 5 1000 100 100 100 100 100 300 300 200 200 200 200 100 200 200 ICO 2640 4500 4500 375010000100001000010000 6220 3750 3750 3750 3750 3750 3750 3750 363168 117650 37120 50490 6 9 1 119 33 TRIS 6 1 119 33 TRIS OA 35 660 30 1 6 1 119 33 TRIS t i •LOGOFF 95 Format of Headers Cards Used before Preliminary Information Cards Preliminary information cards must be preceded by header cards. The cards used and their order are as follows: A card with the number J^ in cc 1; Internal standard retention time cards; A card with the number 2_ in cc 1; Internal standard peak area cards; A card with the number 3^ in cc 1; Cards for factor for amount of filtrate; A card with the number 4_ in cc 1; Cards for factor for filtrate yield; A card with the number 5_ in cc 1; Card(s) for the peak area for the internal standard in the reference experiment. Format of Preliminary Information Cards Instructions for the inclusion of preliminary information are as follows: Internal standard retention time: , one per GLC run (up to n = 500); F 5. 1 fields left to right for n/16 cards; alteration of format statement 171 required if n 4 120; Internal standard peak area: one per GLC run (up to n = 500); F 10. 1 fields left to right for n/8 cards; alteration of format state- ment 172 required if n ^ 120; 96 Factor for amount of filtrate: one per GLC run (up to n = 500); F 5. 3 fields left to right for n/l6 cards; alteration of format statement 173 required if n 4 120; Factor for filtrate yield: one per GLC run (up to n = 500); F 5.4 fields left to right for n/16 cards; alteration of format statement 174 required if n 4 120; Peak area for the internal standard in the reference experi- ment: up to 4 reference areas; F 10. 1 fields left to right. Format of the Card Used before the Group of Cards for the Peaks in a GLC Run Before the group of cards for the peaks in a GLC run, a card with the following information is needed: cc 1 The number 6; cc 3 The number 9; cc 5 The numerical position in the array of internal standard reference areas occupied by the internal standard reference area to be used for this GLC run (1, 2, 3, or 4); cc 11-13 GLC run number (1 to n); cc 16-19 GLC run identification; cc 21-25 Column identification. Format Used for GLC Peak Cards One card is used for each peak in the GLC run. The following information is required: cc 1 The number 6; cc 5 Same as previous card; 97 cc 11-13 GLC run number; cc 16-19 GLC run identification; cc 21-25 Column identification; cc 27-30 Peak identification; cc 32-35 Peak retention time (F 4. 1); cc 37-41 Peak height (F 5. 1); cc 42-46 Peak half-width (F 5. 1); cc 48-51 Attenuation multiplier (F 4.0). Format for Total Peak Area Summation Card For each GLC run a key card causing total peak area for the run to be added follows the last peak card (for run rj and comes before the preliminary card used before.the next GLC run (run (r_ + I)). The following information is required: cc 1 The number 6; cc 5 The numerical position in the array of internal standard reference areas occupied by the internal standard reference area to be used for this GLC run (1, 2, 3, or 4); cc 11-13 GLC run number (1 to n); cc 16-19 GLC run identification; cc 21-25 Column identification. Output of the Program The output for each GLC run occupies three pages of line printer paper. The first page of output for a GLC run gives the following information: 98 Experiment identification; Column identification; Internal standard retention time; Internal standard peak area; Key for the reference area to which the internal standard peak area was compared (1, 2, 3, or 4); Reference area to which the internal standard peak area was compared; Area normalizing factor; Normalizing factor for amount of filtrate; Normalizing factor for filtrate yield; Overall normalizing factor. The second page of output for a GLC run gives the following information in tabular form: Experiment identification; Column identification; Peak identification; Peak retention time; Peak retention time relative to internal standard retention time; Peak height; Peak half-width; Attenuation multiplier; Peak area; Normalized peak area. The third page of output for a GLC run lists the following information: Experiment identification; Column identification; 99 GLC run number (n); Total peak area for the GLC run; Total normalized peak area for the GLC run. 100 APPENDIX II List of Abbreviations This appendix gives the abbreviations used in this thesis which are not in standard use in chemical and biochemical literature. Carbowax Carbowax 20M FD Freeze dried FQM Fair quality moldy FSS Frozen sugared sliced GLC Gas liquid chromatography HS Heads pace IQF Individually quick frozen M Moldy MS Mass spectrometry NGQ Normal good quality OR Overripe TRIS 1,2, 3- tris (2-Cyanoethoxy)- propane UR Underripe VM Very moldy TNPA Total normalized peak area 101 APPENDIX III Illustration of Pooling of Means and Variances for Mold-Level Analysis of Variance Groupings 3x3 Grouping for Peak A IQF Canned FD NGQ X1A^X1B X2A',X2B X3A,X3B FQM X X X X X X 4A; 4B 5A' 5B 6A' 6B M X7A'X7B X8A'X8B X9A,X9B 2x3 Grouping for Peak A Fresh Preserves FSS NGQ X10A'X10B X11A'X11B X12A'X12B M X13A,X13B X14A'X14B X15A'X15B VM IQF X16A'X16B Pooling of Means for Peak A Ex. . + Sx.. IA iB_ x NGQ 12 where i = 1, 2, 3, 10, 11, and 12 102 Ex. „ + 2x.„ lA iB_ XFQM " 6 where i - 4, 5, and 6 2x . + 2x.„ iA LB x M 12 where i = 7, 8, 9, 13, 14, and 15 - X16A + X16B XVM 2 Pooled Variance for Peak A _ ESSa.„ + ESSa._ + Q c2 _ (3x3) (2x3) b " 16 _ _ (X16A ° X16B)2 2 - a ESS = Error sum of squares.